From brian.e.carpenter at gmail.com Mon Nov 1 14:58:26 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Tue, 2 Nov 2021 10:58:26 +1300 Subject: [ih] Fwd: Interviews with the founders of the Internet (Kleinrock, Metcalfe, Perlman, Cerf, Mockapetris) In-Reply-To: References: Message-ID: <51484b7c-c26b-203e-21cb-f8e717f847ca@gmail.com> Forwarded with permission: -------- Forwarded Message -------- Subject: Interviews with the founders of the Internet (Kleinrock, Metcalfe, Perlman, Cerf, Mockapetris) Date: Mon, 1 Nov 2021 19:28:24 +0000 In the expectation that this would be of interest to readers on this mailing list, I?m passing on some email from Prof. George Varghese at UCLA: ?In Fall 2020 during COVID I invited five inventors of the Internet to come to my UCLA network class to be interviewed by undergraduates. Here is a summary video:https://youtu.be/sGHnX8WocHw Len Kleinrock talks about scaling laws in large networks and the importance of picking good problems; Bob Metcalfe talks about the rat's nest of wires that preceded Ethernet and the need for more bandwidth for distributed computing; Radia Perlman describes how the constraints of no station changes and constant overhead led to the Spanning Tree algorithm; Paul Mockapetris talks of his goal to scale DNS to every person in the US and why he did not use a database; finally, Vint Cerf describes how IP's design flowed from its design constraints. It?s half an hour but if you want to pick a few minutes I really like the contrast between the UCLA kids (in dorms during COVID) and the Internet pioneers especiallyat 20:57where Vint Cerf talks about how and why he worked on IP.? with kind regards keshav From internet-history at gtaylor.tnetconsulting.net Mon Nov 1 16:26:15 2021 From: internet-history at gtaylor.tnetconsulting.net (Grant Taylor) Date: Mon, 1 Nov 2021 17:26:15 -0600 Subject: [ih] Fwd: Interviews with the founders of the Internet (Kleinrock, Metcalfe, Perlman, Cerf, Mockapetris) In-Reply-To: <51484b7c-c26b-203e-21cb-f8e717f847ca@gmail.com> References: <51484b7c-c26b-203e-21cb-f8e717f847ca@gmail.com> Message-ID: <9204990b-39ee-1228-9631-244bba3f506c@spamtrap.tnetconsulting.net> On 11/1/21 3:58 PM, Brian E Carpenter via Internet-history wrote: > Forwarded with permission: > > On 11/1/21 1:28 PM, Prof. George Varghese wrote: >> In the expectation that this would be of interest to readers on this >> mailing list, I?m passing on some email from Prof. George Varghese at UCLA: Thank you for forwarding the message. I'll check out the video. I would also be interested in checking out the referenced mailing list. Is said mailing list public? Will you please share some information? -- Grant. . . . unix || die From brian.e.carpenter at gmail.com Mon Nov 1 17:56:55 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Tue, 2 Nov 2021 13:56:55 +1300 Subject: [ih] Fwd: Interviews with the founders of the Internet (Kleinrock, Metcalfe, Perlman, Cerf, Mockapetris) In-Reply-To: <9204990b-39ee-1228-9631-244bba3f506c@spamtrap.tnetconsulting.net> References: <51484b7c-c26b-203e-21cb-f8e717f847ca@gmail.com> <9204990b-39ee-1228-9631-244bba3f506c@spamtrap.tnetconsulting.net> Message-ID: <590d1b71-586b-3bc0-3466-248872e729cf@gmail.com> On 02-Nov-21 12:26, Grant Taylor via Internet-history wrote: > On 11/1/21 3:58 PM, Brian E Carpenter via Internet-history wrote: >> Forwarded with permission: >> >> On 11/1/21 1:28 PM, Prof. George Varghese wrote: >>> In the expectation that this would be of interest to readers on this >>> mailing list, I?m passing on some email from Prof. George Varghese at UCLA: > > Thank you for forwarding the message. I'll check out the video. > > I would also be interested in checking out the referenced mailing list. > Is said mailing list public? Will you please share some information? No, sorry, I should have excised that reference. Keshav and I happen to be on a Cambridge University list. Brian From jeanjour at comcast.net Sun Nov 14 06:22:42 2021 From: jeanjour at comcast.net (John Day) Date: Sun, 14 Nov 2021 09:22:42 -0500 Subject: [ih] Origins of Go-Back-N Message-ID: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> I think this came up before, but I am going to ask anyway: Can someone point me at the first proposal for Go-Back-N? When and by whom? Is it before or after SDLC, which according to the wikipedia is 1975? Thanks, John From jack at 3kitty.org Sun Nov 14 07:59:14 2021 From: jack at 3kitty.org (Jack Haverty) Date: Sun, 14 Nov 2021 15:59:14 +0000 (UTC) Subject: [ih] Origins of Go-Back-N In-Reply-To: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> References: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> Message-ID: My earliest recollection of go-back is it's use in Morse code while transferring messages between 2 stations.? At least that's when I first used it, circa 1964, in amateur radio "traffic nets".?? I used that experience later while working on Arpanet and Internet projects. Jack Haverty From touch at strayalpha.com Sun Nov 14 09:15:00 2021 From: touch at strayalpha.com (touch at strayalpha.com) Date: Sun, 14 Nov 2021 09:15:00 -0800 Subject: [ih] Origins of Go-Back-N In-Reply-To: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> References: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> Message-ID: <47CD4955-6EDA-4353-91E9-104C32BE17EC@strayalpha.com> Hi, John, I don?t know the first instance of go-back-N, but selective repeat (in this case, selective NACK) was used by Edelcrantz in the Swiss countryside optical telegraph in roughly 1800 according to Wikipedia (I?ve seen similar descriptions in a little book the IEEE published on the history of telecom). The details of Chappe?s optical telegraph in France weren?t documented, so may have had similar functions and predate this. I wouldn?t be surprised if go-back-N had origins near that time or before. Joe ? Joe Touch, temporal epistemologist www.strayalpha.com > On Nov 14, 2021, at 6:22 AM, John Day via Internet-history wrote: > > I think this came up before, but I am going to ask anyway: > > Can someone point me at the first proposal for Go-Back-N? When and by whom? > > Is it before or after SDLC, which according to the wikipedia is 1975? > > Thanks, > John > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From bpurvy at gmail.com Sun Nov 14 09:44:42 2021 From: bpurvy at gmail.com (Bob Purvy) Date: Sun, 14 Nov 2021 09:44:42 -0800 Subject: [ih] Origins of Go-Back-N In-Reply-To: <47CD4955-6EDA-4353-91E9-104C32BE17EC@strayalpha.com> References: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> <47CD4955-6EDA-4353-91E9-104C32BE17EC@strayalpha.com> Message-ID: Since we're going back to the pre-Morse era: I'm currently (re)reading *The Count of Monte Cristo*, and you may dimly recall that the Count pays a visit to a telegraph station, in about 1829. He befriends and bribes the telegraph operator, whose job it is to see the signals from his next neighbor and faithfully relay them. He has to stop operating when it's foggy. *Spoiler alert*: the Count bribes him enough to retire, and then has him relay a false signal that causes his nemesis M. Danglars to lose a million francs. This book has been made into a movie at least four times, so I'm eager to see how they treat this episode (or if they do). Yeah, yeah, I'm sure this is all in Wikipedia. I don't want to hear about it ? On Sun, Nov 14, 2021 at 9:15 AM touch--- via Internet-history < internet-history at elists.isoc.org> wrote: > Hi, John, > > I don?t know the first instance of go-back-N, but selective repeat (in > this case, selective NACK) was used by Edelcrantz in the Swiss countryside > optical telegraph in roughly 1800 according to Wikipedia (I?ve seen similar > descriptions in a little book the IEEE published on the history of > telecom). The details of Chappe?s optical telegraph in France weren?t > documented, so may have had similar functions and predate this. > > I wouldn?t be surprised if go-back-N had origins near that time or before. > > Joe > > ? > Joe Touch, temporal epistemologist > www.strayalpha.com > > > On Nov 14, 2021, at 6:22 AM, John Day via Internet-history < > internet-history at elists.isoc.org> wrote: > > > > I think this came up before, but I am going to ask anyway: > > > > Can someone point me at the first proposal for Go-Back-N? When and by > whom? > > > > Is it before or after SDLC, which according to the wikipedia is 1975? > > > > Thanks, > > John > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From jack at 3kitty.org Sun Nov 14 13:07:57 2021 From: jack at 3kitty.org (Jack Haverty) Date: Sun, 14 Nov 2021 13:07:57 -0800 Subject: [ih] Origins of Go-Back-N In-Reply-To: <7454EA0E-4CFF-4F04-A49B-F347710981A6@comcast.net> References: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> <7454EA0E-4CFF-4F04-A49B-F347710981A6@comcast.net> Message-ID: There was a protocol used between two stations to move a message. It was partly formal protocol,? partly informal traditions. A message was similar to an email today but only had one destination.?? Q-codes were used, as well as short interactions between stations.? E.g., a sender might choose to send a long message by sending just a few words at a time and then waiting for a response.?? The receiver might send "R", which meant Received, continue.? Or if there had been interference or staiic, the receiver might send "AA ".?? "AA" meant All After, i.e., he received everything up to the , but please retransmit everything after that. (i.e., "go back").? There were a bunch of such "protocol commands".? E.g., there was some way to negotiate "send only N words at a time".?? Kind of like TCP's Window mechanism??? Both sender and receiver also adapted their behavior depending on conditions and negotiated changes as needed to compensate for signal strengths, interference, etc.? (somewhat analogous to TCP's algorithm's for retransmissions, adaptive routing, et al).? E.g., QRS meant "Slow down", which would help if radio conditions were especially noisy.?? QRQ meant go Quicker - i.e., you're loud and clear you can go faster.?? Modems do much the same thing to move bits as fast as possible but reliably. This was all part of a rather elaborate continent-wide system which created a wide-area radio "internet".?? It was legally limited to US destinations, but included military installations such as GITMO (Guantanamo Bay).? Much of the messages handled were for military service people communicating with their families.?? In the era before cell phones and the Internet, amateur radio was sometimes the only way to communicate. At scheduled times, all stations in a local area would convene on the air at a designated frequency.?? It was similar to a broadcast LAN, except that you couldn't be sure that every station could hear every other station.? Kind of like early Ethernet with no CSMA-CD. To maintain order, one station would act as "Net Control" and manage the activity of all other stations - similar to how some early LANs worked with one node being the manager.? Stations would "check in" to a net using Q-signals.? E.g., I might send "K3FIV QNI QTC 2 MO", which gave my identity (K3FIV), that I was checking in (QNI), and that I had 2 messages to be delivered to Missouri (QTC 2 MO).?? It was somewhat analogous to DHCP and such "initial connection protocols".? There was also a way for a station to indicate that it could "take traffic" for particular destinations, i.e., it was willing to act as the next stage of a message's trip.? Not quite as fancy as SPF but performed a similar function. Net control would collect all that information from all stations, and as it found matches it would send pairs of stations to some other frequency to "pass traffic", i.e., transfer messages using the techniques I described above.?? That was kind of like 3rd-party FTP transfers where two hosts transfer a file at the direction of a third host.?? Such stations would come back to the main net frequency and then report their success or failure. (i.e., ACK or NACK) These "local nets" were geographically restricted by the physics of radio, usually covering perhaps a part of a state. ? So they were part of a larger system to permit "wide area" networking.? There were 9 or so "regional networks" that used different radio bands to cover perhaps a 500-mile radius.? One station from each "local net" would check in to its associated "regional net" to move a message further in the "right direction".?? Such stations are somewhat functionally analogous to the Internet's gateways and routers. Similarly, the "regional nets" were connected by stations that participated in something called, IIRC, "Transcontinental Traffic Corps".?? A message from Massachusetts to California might be handled by 5 or more stations along the way.? The schedules for networks were set in a staggered way so that messages could flow quickly "up the ladder".? But it would typically take another day or so to get back down to the final destination. I was located in EPA - Eastern Pennsylvania.? One or a few stations from the EPA net would be assigned as a "gateway", and subsequently check in to "3RN" which was the 3rd Region Net" covering 3 states: Pennsylvania, Maryland, and Delaware.? Similarly, omeone from 3RN would subsequently check in to the Transcontinental network to get the message to the appropriate regional net for the addressee, and then it would work down to the appropriate local net, and then get delivered, usually by telephone. Sounds a lot like the Internet? For amateur radio, this system began on January 27, 1917.??? See http://www.arrl.org/news/view/transcontinental-relay-recreated-for-100th-anniversary-commemoration Of course as others have noted, such techniques were in use much earlier in Morse over wires in the 19th century.?? I wouldn't be surprised if techniques like this were used also in the days of ships and signal flags or before. ARPA actually had a role in Morse at the same time as the ARPANET was growing -- see https://apps.dtic.mil/sti/pdfs/ADA239925.pdf chapter 22.? I was the "radio expert" we used at MIT in that project to build an "expert system" for handling Morse code by computers in the late 1980s.? We used that "radio traffic network" as the environment in which our computer-based Morse "operator" participated, by making it do in (computer) code what I remembered doing sitting at a Morse key back in the 60s. Shortly after that Morse project at MIT, I joined BBN and my first assignment was implementing TCP for Unix.?? So all of that "networking" experience from amateur radio protocols was still swirling around in my head as we worked on the foundation of the Internet. through the late 70s and early 80s. Hmmm.? I suspect other people involved in the Internet also had prior experience with the radio world, and even Morse.? So in some sense that radio environment is part of Internet history.?? I've always thought that our network architecture, protocols and equipment are based on what people have done in the pre-computer era.?? Computers are of course much faster and don't get bored.? I hope. /Jack Haverty On 11/14/21 10:23 AM, John Day wrote: > Could you elaborate, Jack? How did it work in CW? Was there a Q-code? > > Take care, > John > >> On Nov 14, 2021, at 10:59, Jack Haverty via Internet-history wrote: >> >> My earliest recollection of go-back is it's use in Morse code while transferring messages between 2 stations. At least that's when I first used it, circa 1964, in amateur radio "traffic nets". I used that experience later while working on Arpanet and Internet projects. >> Jack Haverty >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history From brian.e.carpenter at gmail.com Sun Nov 14 13:27:46 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Mon, 15 Nov 2021 10:27:46 +1300 Subject: [ih] Origins of Go-Back-N In-Reply-To: <47CD4955-6EDA-4353-91E9-104C32BE17EC@strayalpha.com> References: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> <47CD4955-6EDA-4353-91E9-104C32BE17EC@strayalpha.com> Message-ID: On 15-Nov-21 06:15, touch--- via Internet-history wrote: > Hi, John, > > I don?t know the first instance of go-back-N, but selective repeat (in this case, selective NACK) was used by Edelcrantz in the Swiss countryside optical telegraph in roughly 1800 according to Wikipedia (I?ve seen similar descriptions in a little book the IEEE published on the history of telecom). The details of Chappe?s optical telegraph in France weren?t documented, so may have had similar functions and predate this. > > I wouldn?t be surprised if go-back-N had origins near that time or before. In human speech, I would guess that "Say again!" or its equivalent has been in use for at least 100,000 years. So Go-Back-N would always have been obvious to one skilled in the art of conversation. More recently, I believe one needs to look at codes like AA in radiotelegraphy, e.g. "AA All after ? (used after a question mark in radiotelegraphy or after RQ in radiotelephony (in case of language difficulties) or after RPT, to request a repetition)." [Rec. ITU-R M.1172] I don't know how old that is. https://en.wikipedia.org/wiki/ARQ-M describes Automatic Repeat reQuest, Multiplex, dated 1947. Brian > > Joe > > ? > Joe Touch, temporal epistemologist > www.strayalpha.com > >> On Nov 14, 2021, at 6:22 AM, John Day via Internet-history wrote: >> >> I think this came up before, but I am going to ask anyway: >> >> Can someone point me at the first proposal for Go-Back-N? When and by whom? >> >> Is it before or after SDLC, which according to the wikipedia is 1975? >> >> Thanks, >> John >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history > From touch at strayalpha.com Sun Nov 14 13:40:04 2021 From: touch at strayalpha.com (touch at strayalpha.com) Date: Sun, 14 Nov 2021 13:40:04 -0800 Subject: [ih] Origins of Go-Back-N In-Reply-To: References: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> <47CD4955-6EDA-4353-91E9-104C32BE17EC@strayalpha.com> Message-ID: <5DE2B706-CF82-4792-9F7E-8AF9683BD318@strayalpha.com> ? Joe Touch, temporal epistemologist www.strayalpha.com > On Nov 14, 2021, at 1:27 PM, Brian E Carpenter wrote: > >> I wouldn?t be surprised if go-back-N had origins near that time > or before. > > In human speech, I would guess that "Say again!" or its equivalent has been in use for at least 100,000 years. So Go-Back-N would always have been obvious to one skilled in the art of conversation. Agreed. CSMA/CD is basically how people speak in a group. IMO, the distinction is also the transition from ?practice? to ?protocol?, i.e., from ?we use it? to ?formal rules of interaction?. Joe From bill.n1vux at gmail.com Sun Nov 14 14:27:34 2021 From: bill.n1vux at gmail.com (Bill Ricker) Date: Sun, 14 Nov 2021 17:27:34 -0500 Subject: [ih] Origins of Go-Back-N In-Reply-To: References: <44EF4BB0-035E-4065-8198-B25E53B4B7C9@comcast.net> <7454EA0E-4CFF-4F04-A49B-F347710981A6@comcast.net> Message-ID: On Sun, Nov 14, 2021 at 4:08 PM Jack Haverty via Internet-history < internet-history at elists.isoc.org> wrote: > There was a protocol used between two stations to move a message. It was > partly formal protocol, partly informal traditions. > > A message was similar to an email today but only had one destination. > Q-codes were used, as well as short interactions between stations. > Change WAS to IS. The ARRL National Traffic System still exists and functions. It's a backup system i hope we never need in the mainland USA, but it remains useful after e.g. Caribean hurricanes. http://www.arrl.org/chapter-one-national-traffic-system interference or staiic, the receiver might send "AA ". "AA" > meant All After, i.e., he received everything up to the , And indeed, it is pronounced ALL AFTER on "phone" (voice) modes. > This was all part of a rather elaborate continent-wide system which > created a wide-area radio "internet". Still operating. Transcontinental Corps (TCC) moves "Traffic" aka Radiogram Messages between Regions 0-9; Region nets feed TCC and route messages to metro areas, and local nets (often on FM voice repeaters but also cloud-warmer HF CW Morse and AX-25 Packet Radio message boards) get the message to a Ham NTS operator in the local phone dialing region of the addressee. TCC and Regions can use both "Original Digital" Morse Code and modern digital modes. (Which is easier than in pre-deregulation days! but local public relations are still good) It was legally limited to US > destinations, but included military installations such as GITMO > (Guantanamo Bay). There is now a list of countries with which State Dept has negotiated Third Party Traffic Agreements. And in disasters, they usually get a short term waiver with other countries to get their Red Cross "I'm ok but the phone's out" messages delivered to stateside relatives. Much of the messages handled were for military > service people communicating with their families. Yes, NTS had good relations with MARS (Military Amateur Radio System) in the old days, and delivered for / routed to MARS stations for soldiers overseas. (MARS stations were run as a hobby-service by soldiers on their personal time.) MARS also provided free phonecalls home via HF AM or SSB to phone-patch (again seeking a Ham volunteer in local dialing area if possible, but SEN Barry Goldwater K7UGA's MARS stations AFA7UGA & AGA6BG, staffed by volunteers at his mansion 7x24, would place the tollcall from there if no one in the right Area Code or local dialing area could be found. In the era before > cell phones and the Internet, amateur radio was sometimes the only way > to communicate. > Yes indeed. And when the infrastructure is down, the Sat Phone or Satellite Internet is the only alternative still. > Sounds a lot like the Internet? > > I was the "radio expert" we used at MIT in that project to build an > If you haven't followed, you'll be relieved to hear that W1XM/W1MX successfully rallied alumni etc to save the Green Bldg Radome (which EE and Meteorology Depts were ready to remove, and would have likely have compromised the W1XM VHF+ station). I was located in EPA - Eastern Pennsylvania. 73 OM de N1VUX EMA -- Bill Ricker bill.n1vux at gmail.com https://www.linkedin.com/in/n1vux From steve at shinkuro.com Mon Nov 15 06:19:31 2021 From: steve at shinkuro.com (Steve Crocker) Date: Mon, 15 Nov 2021 09:19:31 -0500 Subject: [ih] Intel 4004 vs the IMP Message-ID: National Public Radio in the U.S. is noting today that Intel introduced the 4004 microprocessor 50 years ago this week. And thus began the microprocessor revolution. The Arpanet used separate computers, the IMPs, to orchestrate the communications. Financially, this was just barely practical and was a major breakthrough. The IMPs were Honeywell 516 and 316 computers, which used technology several years older than Intel's 4004. And I doubt the 4004 would have been capable of serving as the CPU in an Arpanet router. But more capable microprocessors came along over the next few years. Might it be possible to pin down the crossover date? That is, when did a microprocessor appear that was of the same power as the Honeywell 316? Alternatively, the processing needs for a router weren't staying static, so perhaps matching the power of a H316 would not have been sufficient. In that case, the question is when did a microprocessor appear that was powerful enough to serve as a router? I have a guess, but I don't have first hand knowledge. Comments? Thanks, Steve From jmamodio at gmail.com Mon Nov 15 07:11:19 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 09:11:19 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: I really doubt it, the 4004 and companion chipset were designed for a specific application/customer, the Busicon 141-PF calculator. The event that really pushed intel to get into the microprocessor business was the custom design of the 1201 later transformed into the 8008, based on the original CPU design (TTL based) for the Datapoint 2200 from Computer Terminal Corporation of San Antonio (later renamed as Datapoint,) unfortunately intel was late to deliver and Datapoint never used but made a deal that instead of paying the leftover on the contract intel kept the rights to continue developing and commercializing the microprocessor technology. At first intel was hesitant to take the project, at that time they were the premier manufacturer of memory chips and CTC was one of their major customers, and they didn't want the microprocessor to jeopardize their business with other computer manufacturers at that time. Obviously being San Antonio the birthplace of CTC/Datapoint and the Datapoint 2200, we have rich history collection, for about three years I was involved with the local museum of science and technology where we have a large collection of Datapoint equipment, documents, etc, among them two of the four hand build prototypes of the Datapoint 2200. Also we have first hand knowledge, and close access to people that were closely involved in the company, including a good friend of mine now living in Austin, the son of one of Datapoint's founder. So most of the stories around the 4004 (including the one saying that it was used by NASA for early spacecraft) are pure bologna or popular myths. Cheers Jorge On Mon, Nov 15, 2021 at 8:19 AM Steve Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > National Public Radio in the U.S. is noting today that Intel introduced the > 4004 microprocessor 50 years ago this week. And thus began the > microprocessor revolution. > > The Arpanet used separate computers, the IMPs, to orchestrate the > communications. Financially, this was just barely practical and was a > major breakthrough. The IMPs were Honeywell 516 and 316 computers, which > used technology several years older than Intel's 4004. And I doubt the > 4004 would have been capable of serving as the CPU in an Arpanet router. > But more capable microprocessors came along over the next few years. > > Might it be possible to pin down the crossover date? That is, when did a > microprocessor appear that was of the same power as the Honeywell 316? > > Alternatively, the processing needs for a router weren't staying static, so > perhaps matching the power of a H316 would not have been sufficient. In > that case, the question is when did a microprocessor appear that was > powerful enough to serve as a router? > > I have a guess, but I don't have first hand knowledge. > > Comments? > > Thanks, > > Steve > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From jmamodio at gmail.com Mon Nov 15 07:18:54 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 09:18:54 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: Ohh forgot to mention, the intel 8008 was released in April 1972, but it was 8-bits with a clock up to 800KHz, hard to challenge the HW316 (released in 1969) that was 16-bits with a 2.5MHz clock. Warm, Regards Jorge On Mon, Nov 15, 2021 at 8:19 AM Steve Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > National Public Radio in the U.S. is noting today that Intel introduced the > 4004 microprocessor 50 years ago this week. And thus began the > microprocessor revolution. > > The Arpanet used separate computers, the IMPs, to orchestrate the > communications. Financially, this was just barely practical and was a > major breakthrough. The IMPs were Honeywell 516 and 316 computers, which > used technology several years older than Intel's 4004. And I doubt the > 4004 would have been capable of serving as the CPU in an Arpanet router. > But more capable microprocessors came along over the next few years. > > Might it be possible to pin down the crossover date? That is, when did a > microprocessor appear that was of the same power as the Honeywell 316? > > Alternatively, the processing needs for a router weren't staying static, so > perhaps matching the power of a H316 would not have been sufficient. In > that case, the question is when did a microprocessor appear that was > powerful enough to serve as a router? > > I have a guess, but I don't have first hand knowledge. > > Comments? > > Thanks, > > Steve > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From clemc at ccc.com Mon Nov 15 07:24:19 2021 From: clemc at ccc.com (Clem Cole) Date: Mon, 15 Nov 2021 10:24:19 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: On Mon, Nov 15, 2021 at 9:19 AM Steve Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > Might it be possible to pin down the crossover date? That is, when did a > microprocessor appear that was of the same power as the Honeywell 316? > > Well gioce these are the features of the 216: The programmers' model of the H-316 consisted of the following registers: - The 16-bit *A* register was the primary arithmetic and logic accumulator. - The 16-bit *B* register was used for double-length arithmetic operations. - The 16-bit *program counter* holds the address of the next instruction. - A *carry flag* indicated arithmetic overflow. - A 16-bit *X index* register was also provided for modification of the address of operands. The instruction set had 72 arithmetic, logic, I/O and flow-control instructions. What I don't remember is the clock frequency, but I think it was in the order of .5-1Mz. It used core which teneded to be slower than semiconductor memory at the time. So, I suspect if you just match the ISA's, first commercial microprocessor to come close to that would have been the M6809 which was introduced in 1978, which had two 8-bit A/B accumulators which combined to single 16-bit accumulator but also has a 16-bit D accumulator. It also had 2 16 bit index registers (X and Y). It was usually combined with semiconductor memory and clocked at 2 Mhz. The M68000 would come out as an experimental (unnumberred) chip for a 10 of us a few months later and would be released for GA, in early/mid '79 [I was one of the people with the X-series chip at Tektronix, so I really don't remember the final GA time). Certainly it would been workable. The question is if an 8-bit processor with a 16-bit address space like the 8080/Z80 or 6800/6502 would have been good enough. They all tended to use semiconductor memory, so the memory speed of the 316 is likely to have been able to be matched/exceeded. But the question is open if the code when converted to 8-bit ops to perform what had been done in 16-bits would have been reasonable. My >>WAG<< is that since so many slick video games got built on the 1Mhz 6502, I think an IMP might have been possible but would have taken some very slick and careful coding I suspect. ? From clemc at ccc.com Mon Nov 15 07:26:34 2021 From: clemc at ccc.com (Clem Cole) Date: Mon, 15 Nov 2021 10:26:34 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: Lord I'm more typing challenged today than normal: Well given these are the features of the 316 ? On Mon, Nov 15, 2021 at 10:24 AM Clem Cole wrote: > > > On Mon, Nov 15, 2021 at 9:19 AM Steve Crocker via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Might it be possible to pin down the crossover date? That is, when did a >> microprocessor appear that was of the same power as the Honeywell 316? >> >> Well gioce these are the features of the 216: > > The programmers' model of the H-316 consisted of the following registers: > > > - The 16-bit *A* register was the primary arithmetic and logic > accumulator. > - The 16-bit *B* register was used for double-length arithmetic > operations. > - The 16-bit *program counter* holds the address of the next > instruction. > - A *carry flag* indicated arithmetic overflow. > - A 16-bit *X index* register was also provided for modification of > the address of operands. > > The instruction set had > 72 arithmetic, logic, I/O and flow-control instructions. > > What I don't remember is the clock frequency, but I think it was in the > order of .5-1Mz. It used core which teneded to be slower than > semiconductor memory at the time. > > So, I suspect if you just match the ISA's, first commercial microprocessor > to come close to that would have been the M6809 which was introduced in > 1978, which had two 8-bit A/B accumulators which combined to single 16-bit > accumulator but also has a 16-bit D accumulator. It also had 2 16 bit > index registers (X and Y). It was usually combined with semiconductor > memory and clocked at 2 Mhz. > > The M68000 would come out as an experimental (unnumberred) chip for a 10 > of us a few months later and would be released for GA, in early/mid '79 [I > was one of the people with the X-series chip at Tektronix, so I really > don't remember the final GA time). Certainly it would been workable. > > The question is if an 8-bit processor with a 16-bit address space like the > 8080/Z80 or 6800/6502 would have been good enough. They all tended to use > semiconductor memory, so the memory speed of the 316 is likely to have been > able to be matched/exceeded. But the question is open if the code when > converted to 8-bit ops to perform what had been done in 16-bits would have > been reasonable. > > My >>WAG<< is that since so many slick video games got built on the 1Mhz > 6502, I think an IMP might have been possible but would have taken some > very slick and careful coding I suspect. > > ? > From jmamodio at gmail.com Mon Nov 15 07:40:34 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 09:40:34 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: Just as a time reference the first Proteon router, the p4200 was released in 1986, can't find now what processor they used, same year Cisco released the AGS, in those days the sexiest micro was the Motorola 68000 line. I doubt that anything in the line of the 6500/6800/8080 could match the HW316 or any of the other microcomputer CPU's at the time. Warm Regards Jorge On Mon, Nov 15, 2021 at 9:27 AM Clem Cole via Internet-history < internet-history at elists.isoc.org> wrote: > Lord I'm more typing challenged today than normal: > Well given these are the features of the 316 > > > ? > > On Mon, Nov 15, 2021 at 10:24 AM Clem Cole wrote: > > > > > > > On Mon, Nov 15, 2021 at 9:19 AM Steve Crocker via Internet-history < > > internet-history at elists.isoc.org> wrote: > > > >> Might it be possible to pin down the crossover date? That is, when did > a > >> microprocessor appear that was of the same power as the Honeywell 316? > >> > >> Well gioce these are the features of the 216: > > > > The programmers' model of the H-316 consisted of the following registers: > > > > > > - The 16-bit *A* register was the primary arithmetic and logic > > accumulator. > > - The 16-bit *B* register was used for double-length arithmetic > > operations. > > - The 16-bit *program counter* holds the address of the next > > instruction. > > - A *carry flag* indicated arithmetic overflow. > > - A 16-bit *X index* register was also provided for modification of > > the address of operands. > > > > The instruction set had > > 72 arithmetic, logic, I/O and flow-control instructions. > > > > What I don't remember is the clock frequency, but I think it was in the > > order of .5-1Mz. It used core which teneded to be slower than > > semiconductor memory at the time. > > > > So, I suspect if you just match the ISA's, first commercial > microprocessor > > to come close to that would have been the M6809 which was introduced in > > 1978, which had two 8-bit A/B accumulators which combined to single > 16-bit > > accumulator but also has a 16-bit D accumulator. It also had 2 16 bit > > index registers (X and Y). It was usually combined with semiconductor > > memory and clocked at 2 Mhz. > > > > The M68000 would come out as an experimental (unnumberred) chip for a 10 > > of us a few months later and would be released for GA, in early/mid '79 > [I > > was one of the people with the X-series chip at Tektronix, so I really > > don't remember the final GA time). Certainly it would been workable. > > > > The question is if an 8-bit processor with a 16-bit address space like > the > > 8080/Z80 or 6800/6502 would have been good enough. They all tended to > use > > semiconductor memory, so the memory speed of the 316 is likely to have > been > > able to be matched/exceeded. But the question is open if the code when > > converted to 8-bit ops to perform what had been done in 16-bits would > have > > been reasonable. > > > > My >>WAG<< is that since so many slick video games got built on the 1Mhz > > 6502, I think an IMP might have been possible but would have taken some > > very slick and careful coding I suspect. > > > > ? > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From tony.li at tony.li Mon Nov 15 07:48:20 2021 From: tony.li at tony.li (Tony Li) Date: Mon, 15 Nov 2021 07:48:20 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: <4334078A-EDEA-4219-8DAC-837DA77645F5@tony.li> Steve, > On Nov 15, 2021, at 6:19 AM, Steve Crocker via Internet-history wrote: > > Alternatively, the processing needs for a router weren't staying static, so > perhaps matching the power of a H316 would not have been sufficient. In > that case, the question is when did a microprocessor appear that was > powerful enough to serve as a router? We?re still waiting for that day. No matter what processor (or the amount of memory) that we can convince HW designers to include, it quickly proves inadequate. Our willingness to put anything and everything into BGP makes for insatiable demand. Regards, Tony From clemc at ccc.com Mon Nov 15 08:00:53 2021 From: clemc at ccc.com (Clem Cole) Date: Mon, 15 Nov 2021 11:00:53 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: On Mon, Nov 15, 2021 at 10:41 AM Jorge Amodio wrote: > > Just as a time reference the first Proteon router, the p4200 was released > in 1986, can't find now what processor they used, same year Cisco released > the AGS, in those days the sexiest micro was the Motorola 68000 line. > Yep ... a 10MHz 68000 would have been fine. The 386 proved sufficient for a number of cheap routers shortly thereafter also, in fact, Proteon made one using a cheap wintel frame and putting ISA cards into it [I used a one of these at one point]. > > I doubt that anything in the line of the 6500/6800/8080 could match the > HW316 or any of the other microcomputer CPU's at the time. > Maybe, and part of my gut reaction is to agree with you, and then I think about what Atari, Nintendo and the like did with the 6502. Someone like Dave Hayne would be a better judge I suspect. He and his peeps did some really amazing stuff with those early processors in those days. Hey, I love the 6502, it one of my favorite ISA's and had a ball with it in the mid-1970s programming it at CMU, but I would not have wanted to try it for an IMP replacement. Hey, we used LSI-11's at CMU for the distributed front-end in '76/'77,. Version two (after I left) used a 8086 in a Multbus in 1979/80, and that version was in many ways the prototype that Andy took to Stanford for what would become the Stanford University Network Terminal. The SUN Terminal, of course, with it's multibus chassis is what Cisco used for the AGS a couple of years later. So I think we all agree, but the time of the full 16-processors like the 8086/Z8000/68000, running on semiconductor memory, and IMP like replacement was quite feasible. But having looked at some of the IMP code that has leaked out, and again the existence proof of the game controllers made using a 6502, I have a general belief that if someone has wanted to do it and *had the right motivation*, folks like Dave and co, would have found a way. That motivation is still also an open question. The game control folks were driven to keep costs as low as possible, while offering acceptable/good performance. Something like an IMP and later a router, less so. Performance, I would have expected to be the high order bit and cost would have been important, but people would pay for it. Hey, the early AGC's from Cisco were many, many thousands. ? From jnc at mercury.lcs.mit.edu Mon Nov 15 08:20:42 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Mon, 15 Nov 2021 11:20:42 -0500 (EST) Subject: [ih] Intel 4004 vs the IMP Message-ID: <20211115162042.21BA618C090@mercury.lcs.mit.edu> > From: Steve Crocker > the processing needs for a router weren't staying static ... when did a > microprocessor appear that was powerful enough to serve as a router? There are two things one has to look at in evaluating these early machines as routers: speed and address space. There was an additional complication because, as you note, the requirement (along both axes) were not static - although initially they were pretty low. I don't know _how_ low the speed axis was, because for the first generation of routers, the attached networks (ARPANET, PRNET, etc) were pretty slow, and so were the hosts, so IIRC 'packets/second' numbers were not high on our lists of concerns. The routers worked, we were up to there in other issues, so we didn't focus on speed. (Although I recall being pretty happy when the MOS-based router I did at MIT out-performed the ELF-based router done by BBN.) The very first router was the BBN one on a PDP-11/40, which although a minicomputer, not a micro, is a useful benchmark. The -11/40 is about 1 MIP (for register-register moves; see below) in processing performance, and the OS used (ELF) did allow use of the memory management hardware, so it wasn't limited to the 56KB of the 'basic' PDP-11. The first routers at MIT (and also router-like devices at SRI, and probably also Fuzzballs) were LSI-11's. Although technically not single-chip microcomputers, they were effectively micros (4 VLSI chips on a smallish card with some support circuitry). Their speed is hard to evaluate, since the PDP-11 supported some complex, but slower, addressing modes. E.g. a register-register MOV took 1.7 microseconds, but a memory-memory move (with the addresses not in registers) seems to have taken 1.7+4.9=6.6 usec, so about 1/5th of a MIP - and, like I said, that was enough 'oomph' for the early routers. If you look at the earliest 'true' micro, the 4004, it was never (that I know of) used in a router, although its speed (a basic CPU clock of about 11 usec, with instruction execution timea of 1 or 2 machine cycles) was probably enough, compared to the slightly faster LSI-11 (whose most baroque memory reference modes could take up to 7 usec, on top of the basic 1.7 of the MOV). Its memory space, however, would have been an issue. Although it had a 12-bit address bus, allowing it to directly address 4K bytes of ROM, it could only address 640 bytes of RAM, not really enough. The 8008 was on the borderline of feasible. Although it was a little slower than the 4004, (.04 to .08 MIPS, at a CPU clock 800 KHz - it could be as slow as 200KHz), its 14-bit address address bus allowed access to 16KB - just barely enough, if it didn't need e.g. EGP (which came along later). The 8080 was close to an LSI-11, so I'm pretty sure it would have been doable. I quickly switched to the PDP-11/23 when it became available, and hacked the code to keep packet buffers in high memory (mostly - some DMA interfaces like the 3MBit Ethernet could only use the bottom 64KB), so the entire low 56KB was available to hold code (e.g. EGP) and all other data. Noel From agmalis at gmail.com Mon Nov 15 08:24:53 2021 From: agmalis at gmail.com (Andrew G. Malis) Date: Mon, 15 Nov 2021 11:24:53 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: Jorge et al, I have the possibly unique experience of having written 316/516 IMP code at BBN (in assembler) while also, at MITRE, having written the code for an 8-bit 6502-based layer-2 Ethernet predecessor that used CSMA/CD on a broadband local-area cable infrastructure (also in assembler) called MITREnet. THe MITREnet units had a 6502 processor, a 2K!!!!! PROM for the program memory, and I can't remember how much RAM for packet buffers and variable memory, but it was probably 4K bytes. We had a computer interface, cable radio interface, store and forward with queuing for the packets, and of course CSMA/CD algorithms for cable transmission management. So you could do an awful lot with a 6502 and a 2K PROM! While it wasn't up to snuff with a 316/516 in terms of processing capacity, the 6502 would have been sufficient for a very early router that handled packet ingress, egress, queuing, and a simple routing protocol such as RIP. After I left MITRE, they upgraded the MITRENET units in the early 80s to use the Zilog Z8000, 32K of PROM for program memory, and 64K of RAM. The upgraded units included TCP/IP, FTP, and TELNET implementations as well as packet forwarding and high-speed (for the time) RS-422 DMA-based host interfaces for PDP-11s and VAXes, so it would certainly have had enough processing capacity to be a full router as well. Cheers, Andy On Mon, Nov 15, 2021 at 10:41 AM Jorge Amodio via Internet-history < internet-history at elists.isoc.org> wrote: > Just as a time reference the first Proteon router, the p4200 was released > in 1986, can't find now what processor they used, same year Cisco released > the AGS, in those days the sexiest micro was the Motorola 68000 line. > > I doubt that anything in the line of the 6500/6800/8080 could match the > HW316 or any of the other microcomputer CPU's at the time. > > Warm Regards > Jorge > > > On Mon, Nov 15, 2021 at 9:27 AM Clem Cole via Internet-history < > internet-history at elists.isoc.org> wrote: > > > Lord I'm more typing challenged today than normal: > > Well given these are the features of the 316 > > > > > > ? > > > > On Mon, Nov 15, 2021 at 10:24 AM Clem Cole wrote: > > > > > > > > > > > On Mon, Nov 15, 2021 at 9:19 AM Steve Crocker via Internet-history < > > > internet-history at elists.isoc.org> wrote: > > > > > >> Might it be possible to pin down the crossover date? That is, when > did > > a > > >> microprocessor appear that was of the same power as the Honeywell 316? > > >> > > >> Well gioce these are the features of the 216: > > > > > > The programmers' model of the H-316 consisted of the following > registers: > > > > > > > > > - The 16-bit *A* register was the primary arithmetic and logic > > > accumulator. > > > - The 16-bit *B* register was used for double-length arithmetic > > > operations. > > > - The 16-bit *program counter* holds the address of the next > > > instruction. > > > - A *carry flag* indicated arithmetic overflow. > > > - A 16-bit *X index* register was also provided for modification > of > > > the address of operands. > > > > > > The instruction set > had > > > 72 arithmetic, logic, I/O and flow-control instructions. > > > > > > What I don't remember is the clock frequency, but I think it was in the > > > order of .5-1Mz. It used core which teneded to be slower than > > > semiconductor memory at the time. > > > > > > So, I suspect if you just match the ISA's, first commercial > > microprocessor > > > to come close to that would have been the M6809 which was introduced in > > > 1978, which had two 8-bit A/B accumulators which combined to single > > 16-bit > > > accumulator but also has a 16-bit D accumulator. It also had 2 16 bit > > > index registers (X and Y). It was usually combined with semiconductor > > > memory and clocked at 2 Mhz. > > > > > > The M68000 would come out as an experimental (unnumberred) chip for a > 10 > > > of us a few months later and would be released for GA, in early/mid '79 > > [I > > > was one of the people with the X-series chip at Tektronix, so I really > > > don't remember the final GA time). Certainly it would been workable. > > > > > > The question is if an 8-bit processor with a 16-bit address space like > > the > > > 8080/Z80 or 6800/6502 would have been good enough. They all tended to > > use > > > semiconductor memory, so the memory speed of the 316 is likely to have > > been > > > able to be matched/exceeded. But the question is open if the code when > > > converted to 8-bit ops to perform what had been done in 16-bits would > > have > > > been reasonable. > > > > > > My >>WAG<< is that since so many slick video games got built on the > 1Mhz > > > 6502, I think an IMP might have been possible but would have taken > some > > > very slick and careful coding I suspect. > > > > > > ? > > > > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From touch at strayalpha.com Mon Nov 15 08:49:19 2021 From: touch at strayalpha.com (touch at strayalpha.com) Date: Mon, 15 Nov 2021 08:49:19 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: <23F2956C-0C8B-4E2F-BE06-620E408EFAF8@strayalpha.com> On Nov 15, 2021, at 7:24 AM, Clem Cole via Internet-history wrote: > > So, I suspect if you just match the ISA's, first commercial microprocessor > to come close to that would have been the M6809 which was introduced in > 1978, which had two 8-bit A/B accumulators which combined to single 16-bit > accumulator but also has a 16-bit D accumulator. It also had 2 16 bit > index registers (X and Y). It was usually combined with semiconductor > memory and clocked at 2 Mhz. The challenge with the 6809 was the complexity of the memory interfacing circuitry; it was only after Motorola developed the 6883 SAM chip that the pair took root, notably as the core of the Radio Shack Color Computer. That happened sometime between 1978 and 1980, the latter being the introduction of the TRS VideoTex terminal and CoCo, which were basically the same box with slight variation (the former had an internal modem, e.g.). The 6809 does come close to the H-316: >> The programmers' model of the H-316 consisted of the following registers: >> >> - The 16-bit *A* register was the primary arithmetic and logic >> accumulator. >> - The 16-bit *B* register was used for double-length arithmetic >> operations. 6809 had one 16-bit D register. >> - The 16-bit *program counter* holds the address of the next >> instruction. Same. >> - A *carry flag* indicated arithmetic overflow. Same. >> - A 16-bit *X index* register was also provided for modification of >> the address of operands. The 6809 had two 16-bit index registers (X, Y) *and* two 16-bit stack pointers (U, S), in addition to the PC above, and a 64KB native address space (this was increased using bank switching in some motherboards). Notably, also, it has the test instruction first appearing IRL in the 6800 that gave us ?halt and catch fire?. The 6809 was far ahead of its time IMO; it supported 100% position independent code and although having fewer instructions than some others (59), the instructions were highly orthogonal. Given its clock (just under 2 MHz) and performance (2x that of a 6502 or Z80), I would interesting to see if it could have kept pace with the H-316. Side note: I learned assembler programming the 6908 in a Coco by hand-assembling code and using BASIC POKE instructions, even to the point of creating new instructions and patching them into the Coco ROMs (after being copied into RAM), using roadmaps meticulously dissected and described in BYTE magazine. I also used the Coco as a terminal into a Xerox SIGMA/6 in college back in 1981, using an acoustic modem at 300 baud to write SNOBOL and Pascal from my dorm room. But I?m feeling much better now ;-) Joe From jnc at mercury.lcs.mit.edu Mon Nov 15 08:59:40 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Mon, 15 Nov 2021 11:59:40 -0500 (EST) Subject: [ih] Intel 4004 vs the IMP Message-ID: <20211115165940.42D7718C090@mercury.lcs.mit.edu> > From: Jorge Amodio > Just as a time reference the first Proteon router, the p4200 was > released in 1986, can't find now what processor they used It was an MC68K, but that was _MUCH_ later; the BBN router (called 'gateway' then) was operating almost a _decade_ earlier, in 1977. > From: Clem Cole > in fact, Proteon made one using a cheap wintel frame and putting ISA > cards into it Yeah, the 4100, after the Multibus-based 4200, as a cheaper but slower alternative. I forget what kind of CPU it had; I think maybe a 386, but it might have been a 286. I'm trying to remember what kind of Intel machine I first brought the CGW up on; I think it was a 286, because I distinctly recall an acerbic exchange with the people at Kaypro (I had a Kaypro 286 machine - my first 386 was an HP), I wanted to use their BIOS for console I/O, with my C version of MOS as the basic OS, and I was irked that unlike IBM, they didn't make their BIOS code (actually, they licensed someone else's) openly available. I threatened that I'd disassemble the code, and release it! In the end I figured out how to used BIOS calls to do it. Noel From winowicki at yahoo.com Mon Nov 15 09:30:42 2021 From: winowicki at yahoo.com (Bill Nowicki) Date: Mon, 15 Nov 2021 17:30:42 +0000 (UTC) Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <20211115165940.42D7718C090@mercury.lcs.mit.edu> References: <20211115165940.42D7718C090@mercury.lcs.mit.edu> Message-ID: <1054318246.862145.1636997442674@mail.yahoo.com> My perspective: the routers based on the Stanford University Network modules that Andy Bechtolscheim designed in 1980 (which evolved into the Cisco router) used the Motorola 68000, and before that was the work Noel did at MIT, and of course the "fuzz balls" of David Mills at U Del on the LSI-11. So late 1970s is probably the time you are looking for. Bill On Monday, November 15, 2021, 08:59:45 AM PST, Noel Chiappa via Internet-history wrote: ? ? > From: Jorge Amodio ? ? > Just as a time reference the first Proteon router, the p4200 was ? ? > released in 1986, can't find now what processor they used It was an MC68K, but that was _MUCH_ later; the BBN router (called 'gateway' then) was operating almost a _decade_ earlier, in 1977. ? ? > From: Clem Cole ? ? > in fact, Proteon made one using a cheap wintel frame and putting ISA ? ? > cards into it Yeah, the 4100, after the Multibus-based 4200, as a cheaper but slower alternative. I forget what kind of CPU it had; I think maybe a 386, but it might have been a 286. I'm trying to remember what kind of Intel machine I first brought the CGW up on; I think it was a 286, because I distinctly recall an acerbic exchange with the people at Kaypro (I had a Kaypro 286 machine - my first 386 was an HP), I wanted to use their BIOS for console I/O, with my C version of MOS as the basic OS, and I was irked that unlike IBM, they didn't make their BIOS code (actually, they licensed someone else's) openly available. I threatened that I'd disassemble the code, and release it! In the end I figured out how to used BIOS calls to do it. ? Noel -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From brez at brezworks.com Mon Nov 15 10:01:30 2021 From: brez at brezworks.com (Jeremy Bresley) Date: Mon, 15 Nov 2021 13:01:30 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <1054318246.862145.1636997442674@mail.yahoo.com> References: <20211115165940.42D7718C090@mercury.lcs.mit.edu> <1054318246.862145.1636997442674@mail.yahoo.com> Message-ID: <0c0396b6-a61c-a932-d63e-3c9a0071f9a0@brezworks.com> On 11/15/21 12:30, Bill Nowicki via Internet-history wrote: > My perspective: the routers based on the Stanford University Network modules that Andy Bechtolscheim designed in 1980 (which evolved into the Cisco router) used the Motorola 68000, and before that was the work Noel did at MIT, and of course the "fuzz balls" of David Mills at U Del on the LSI-11. So late 1970s is probably the time you are looking for. > Bill On Monday, November 15, 2021, 08:59:45 AM PST, Noel Chiappa via Internet-history wrote: > > ? ? > From: Jorge Amodio > > ? ? > Just as a time reference the first Proteon router, the p4200 was > ? ? > released in 1986, can't find now what processor they used > > It was an MC68K, but that was _MUCH_ later; the BBN router (called 'gateway' > then) was operating almost a _decade_ earlier, in 1977. > > ? ? > From: Clem Cole > > ? ? > in fact, Proteon made one using a cheap wintel frame and putting ISA > ? ? > cards into it > > Yeah, the 4100, after the Multibus-based 4200, as a cheaper but slower > alternative. I forget what kind of CPU it had; I think maybe a 386, but it > might have been a 286. Just to fill in some info on the slightly later router history. The Cisco AGS was released in 1986 and used the M68000 CPU connected to a Multibus backplane.? Lots more details available here: https://www.ciscopress.com/articles/article.asp?p=25296 The later AGS+ models kept the same M68000 CPU but had an improved bus (Cbus) for 5 of the slots to allow for higher throughput. The 2500 series that shipped in 1993 (as well as others like the AS5200s based on them) used a Motorola M68360 CPU. Jeremy From dhc at dcrocker.net Mon Nov 15 10:19:00 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Mon, 15 Nov 2021 10:19:00 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: On 11/15/2021 6:19 AM, Steve Crocker via Internet-history wrote: > Might it be possible to pin down the crossover date? That is, when did a > microprocessor appear that was of the same power as the Honeywell 316? > > Alternatively, the processing needs for a router weren't staying static, so > perhaps matching the power of a H316 would not have been sufficient. In > that case, the question is when did a microprocessor appear that was > powerful enough to serve as a router? Shortly after the HP handheld calculator came out, one of the engineers who worked on it did a short article comparing that device to ENIAC, built 25 years earlier. It had a section comparing them as computers, and then a section comparing things like cost, size, reliability, power, etc. The 'computing' part of the exercise showed them to be remarkably similar. The differences in the other section were impressive, of course. Hand vs. room. MTBF of hours versus years. Etc. I think your initial question could nicely translate into a similar exercise, with the constraint that the 'computing' part is required to be roughly the same. Your alternative exercise, is separately worth doing, since there were entirely adequate microprocessor-based routers by the latter 1980s. That one might need to distinguish between 'enterprise' routers vs. 'backbone' routers is worth keeping in mind, though. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From craig at tereschau.net Mon Nov 15 10:39:47 2021 From: craig at tereschau.net (Craig Partridge) Date: Mon, 15 Nov 2021 11:39:47 -0700 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: On Mon, Nov 15, 2021 at 11:19 AM Dave Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > > > I think your initial question could nicely translate into a similar > exercise, with the constraint that the 'computing' part is required to > be roughly the same. > > Your alternative exercise, is separately worth doing, since there were > entirely adequate microprocessor-based routers by the latter 1980s. > That one might need to distinguish between 'enterprise' routers vs. > 'backbone' routers is worth keeping in mind, though. > > Hidden in here is the fact that the algorithms in routers evolved over time. But for the route lookup, the IP part of forwarding is really simple and has a high degree of concurrency (e.g. you can do all the checks for format validity in parallel and then check if any of them failed at one point). That wasn't terribly useful when processors did single instructions, but by the early 1990s, some could do two instructions or more per clock tick, it was a win. As a result, by the early 1990s, most of the instructions executed to forward an IP packet were actually in the device driver (and the layer 2 stuff the driver didn't do that you had to do in software). Route lookup was also an area of tremendous innovation. Key was splitting the forwarding table/FIB from the master routing table. It both allowed efficient update of the forwarding table and made it possible to structure the forwarding table for lookup, without worrying about how it was structured for updating (whereas the master routing table needed to efficiently update). The result was route lookups became an O(log n) or better operation. Craig Craig Partridge's email account for professional society activities and mailing lists. From jmamodio at gmail.com Mon Nov 15 11:56:40 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 13:56:40 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: Not sure what HP calculator they were looking at but ENIAC as far as I know was hardly programmable and it was more time down on maintenance than running, perhaps UNIVAC 1? Also for the comparison I'd use off the shelf micros, there were not that many in the early 70's. I remember the BBN working on a PDP-11 based router and later when Bob Hinden joined they got one done with the LSI-11, but that was around 1981-1983. In the mid 80's and the era of Interop a lot of stuff started to show up.I remember Excelan using 80186 to run TCP/IP on ISA boards, Cheers Jorge On Mon, Nov 15, 2021 at 12:19 PM Dave Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > On 11/15/2021 6:19 AM, Steve Crocker via Internet-history wrote: > > Might it be possible to pin down the crossover date? That is, when did a > > microprocessor appear that was of the same power as the Honeywell 316? > > > > Alternatively, the processing needs for a router weren't staying static, > so > > perhaps matching the power of a H316 would not have been sufficient. In > > that case, the question is when did a microprocessor appear that was > > powerful enough to serve as a router? > > > Shortly after the HP handheld calculator came out, one of the engineers > who worked on it did a short article comparing that device to ENIAC, > built 25 years earlier. It had a section comparing them as computers, > and then a section comparing things like cost, size, reliability, power, > etc. > > The 'computing' part of the exercise showed them to be remarkably > similar. The differences in the other section were impressive, of > course. Hand vs. room. MTBF of hours versus years. Etc. > > I think your initial question could nicely translate into a similar > exercise, with the constraint that the 'computing' part is required to > be roughly the same. > > Your alternative exercise, is separately worth doing, since there were > entirely adequate microprocessor-based routers by the latter 1980s. > That one might need to distinguish between 'enterprise' routers vs. > 'backbone' routers is worth keeping in mind, though. > > d/ > -- > Dave Crocker > Brandenburg InternetWorking > bbiw.net > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From johnl at iecc.com Mon Nov 15 12:08:48 2021 From: johnl at iecc.com (John Levine) Date: 15 Nov 2021 15:08:48 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: Message-ID: <20211115200850.0D0D12FF8B69@ary.qy> It appears that Clem Cole via Internet-history said: >The instruction set had 72 >arithmetic, logic, I/O and flow-control instructions. > >What I don't remember is the clock frequency, but I think it was in the >order of .5-1Mz. It used core which teneded to be slower than >semiconductor memory at the time. Here's a handy programmer's reference that tells us that the memory cycle time, which tended to be the limit on performance in that era, was 1.6us on the 316 and 960ns on the 516: http://bitsavers.org/pdf/computerControlCompany/series16/h316/70130072156_316_516_PgmrRef_Nov70.pdf I was under the impression that they chose the 316 because it had a multi-level priority interrupt which was unusual at the time. There were a lot of 16 bit minis available in 1969 such as the HP2100, DG Nova, and Lockheed Mac16, and the 316 was otherwise quite a me-too machine. R's, John From amckenzie3 at yahoo.com Mon Nov 15 12:21:47 2021 From: amckenzie3 at yahoo.com (Alex McKenzie) Date: Mon, 15 Nov 2021 20:21:47 +0000 (UTC) Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <20211115200850.0D0D12FF8B69@ary.qy> References: <20211115200850.0D0D12FF8B69@ary.qy> Message-ID: <2013012847.660096.1637007707394@mail.yahoo.com> BBN chose the 516 because of its interrupt structure (as you remember) and because it came in a ruggedized cabinet.? We chose the 316 as a follow-on version because it was program-compatible and cheaper.? The fact that the manufacturer was close to BBN and willing to build interfaces to BBN's design (and provide field service for them) was also a factor. Cheers,Alex On Monday, November 15, 2021, 03:09:02 PM EST, John Levine via Internet-history wrote: It appears that Clem Cole via Internet-history said: >The instruction set had 72 >arithmetic, logic, I/O and flow-control instructions. > >What I don't remember is the clock frequency, but I think it was in the >order of .5-1Mz.? It used core which teneded to be slower than >semiconductor memory at the time. Here's a handy programmer's reference that tells us that the memory cycle time, which tended to be the limit on performance in that era, was 1.6us on the 316 and 960ns on the 516: http://bitsavers.org/pdf/computerControlCompany/series16/h316/70130072156_316_516_PgmrRef_Nov70.pdf I was under the impression that they chose the 316 because it had a multi-level priority interrupt which was unusual at the time.? There were a lot of 16 bit minis available in 1969 such as the HP2100, DG Nova, and Lockheed Mac16, and the 316 was otherwise quite a me-too machine. R's, John -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From johnl at iecc.com Mon Nov 15 12:26:26 2021 From: johnl at iecc.com (John Levine) Date: 15 Nov 2021 15:26:26 -0500 Subject: [ih] an excursion into the ENIAC, Intel 4004 vs the IMP In-Reply-To: Message-ID: <20211115202627.3E97C2FF8E27@ary.qy> It appears that Jorge Amodio via Internet-history said: >Not sure what HP calculator they were looking at but ENIAC as far as I know >was hardly programmable and it was more time down on maintenance than >running, perhaps UNIVAC 1? In the late 1940s when they realized that the IAS machine was making slow progress, Adele Goldstine, Von Neuman and others figured out a way to wire ENIAC as a simple stored program computer. It ran Monte Carlo simulations of nuclear fission triggers for H bombs, with the code written by his wife Klari. After 1948 they had high reliability tubes so it could run for a few days between failures, and find and replaace a failed unit in 15 minutes. It was used until 1955 and they even added a small core memory in 1953. There's an article about it in the April 2014 Annals of the History of Computing. obInternet: Dave Walden's "The Arpanet IMP Progam: Retrospective and Ressurection" is in the same issue. R's, John From jack at 3kitty.org Mon Nov 15 12:33:48 2021 From: jack at 3kitty.org (Jack Haverty) Date: Mon, 15 Nov 2021 12:33:48 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <2013012847.660096.1637007707394@mail.yahoo.com> References: <20211115200850.0D0D12FF8B69@ary.qy> <2013012847.660096.1637007707394@mail.yahoo.com> Message-ID: IIRC, another of the important criteria for selecting the Honeywell 516 was the I/O performance characteristics, since much of what the IMP did was transmit and receive over multiple interfaces (to hosts as well as modems) simultaneously.?? The interrupt scheme was part of the larger issue of I/O capability. So in looking for the earliest "comparable" microprocessor, in addition to comparing metrics such as CPU speed and memory, I think you have to look at I/O characteristics, and how good the processor was at moving data in and out of multiple serial interfaces simultaneously. /Jack Haverty On 11/15/21 12:21 PM, Alex McKenzie via Internet-history wrote: > BBN chose the 516 because of its interrupt structure (as you remember) and because it came in a ruggedized cabinet.? We chose the 316 as a follow-on version because it was program-compatible and cheaper.? The fact that the manufacturer was close to BBN and willing to build interfaces to BBN's design (and provide field service for them) was also a factor. > > Cheers,Alex > > On Monday, November 15, 2021, 03:09:02 PM EST, John Levine via Internet-history wrote: > > It appears that Clem Cole via Internet-history said: >> The instruction set had 72 >> arithmetic, logic, I/O and flow-control instructions. >> >> What I don't remember is the clock frequency, but I think it was in the >> order of .5-1Mz.? It used core which teneded to be slower than >> semiconductor memory at the time. > Here's a handy programmer's reference that tells us that the memory cycle time, which tended > to be the limit on performance in that era, was 1.6us on the 316 and 960ns on the 516: > > http://bitsavers.org/pdf/computerControlCompany/series16/h316/70130072156_316_516_PgmrRef_Nov70.pdf > > I was under the impression that they chose the 316 because it had a multi-level priority interrupt > which was unusual at the time.? There were a lot of 16 bit minis available in 1969 such as the HP2100, > DG Nova, and Lockheed Mac16, and the 316 was otherwise quite a me-too machine. > > R's, > John From jnc at mercury.lcs.mit.edu Mon Nov 15 13:53:29 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Mon, 15 Nov 2021 16:53:29 -0500 (EST) Subject: [ih] Intel 4004 vs the IMP Message-ID: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> > From: Jack Haverty > IIRC, another of the important criteria for selecting the Honeywell 516 > was the I/O performance characteristics > ... > So in looking for the earliest "comparable" microprocessor, in addition > to comparing metrics such as CPU speed and memory, I think you have to > look at I/O characteristics Yes, but... in a _router_, the CPU didn't need to _ever_ look at most of the data in the packet. In anything that did TCP, yeah, you had to do the checksum, and that almost always needed the CPU to fondle each byte, but in e.g. the CGW, since it _never_ copied packets around (not even e.g. to make room for a longet physical network header on the front), if a packet came in one DMA interface, and out another, the CPU never even saw most of thwe bytes in the packet, so the CPU speed was not too relevant, it was all bus-bandwidth dependant. Early on, not all network interfaces were DMA; there were three different approaches: - DMA - full packet buffers in the interface, but the CPU had to manually move bytes from the interface to buffers in memory, so 3 bus cycles/word (with unrolled loop with pointers to device and buffer in registers i) instruction fetch, ii) read from interface, iii) write to memory) so 3 times as much bus traffic per word, compared to DMA - interrupt per word But even the latter wasn't _necessarily_ a problem; at MIT, the main ARPANET gateway for quite a while used the Stanford/SRI 1822 Interface: https://gunkies.org/wiki/Stanford_1822_Interface which was interrupt/byte, but peformance wasn't a problem (that I recall). Like I said, for all early routers, performance was not an issue. Performance only became an issue when there were multiple COTS router vendors, when performance became an easy way for their marketing people to distinguish their products from those of competitors. I doubt the users could ever have told. I don't know if _early_ microprocessors (i.e. long before Motorola 68Ks, Intel x86's, etc) supported DMA on theit memory busses; even for those that didn't, it might have been possible to build an external bus controller, and either stall the CPU (if the memory bus was busy doing DMA), or build a multi-port main memory, or something. Noel From dhc at dcrocker.net Mon Nov 15 14:34:22 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Mon, 15 Nov 2021 14:34:22 -0800 Subject: [ih] Museum-quality archive for this list? Message-ID: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> I don't remember whether this has been discussed or resolved here, before, but looking at the depth of historical knowledge that is casually proffered here makes me wish that the list is automatically archived under museum-quality conditions, with easy public access. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From jack at 3kitty.org Mon Nov 15 14:54:16 2021 From: jack at 3kitty.org (Jack Haverty) Date: Mon, 15 Nov 2021 14:54:16 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> Message-ID: <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> True for more modern systems, but in the era of the 316/516, inexpensive computers sometimes did I/O in the simplest possible manner.? E.g., to handle I/O on a serial interface, the CPU might have to take an interrupt on every byte, read that byte from the hardware interface, and re-enable the interface quickly enough for it to be ready to handle the next byte,?? The hardware only buffered a single byte at a time.? The CPU also had to do that all fast enough to handle all the streams of interrupts from all its interfaces in order not to lose data. ? That would occur if a particular line received a byte and raised its interrupt, but the processor was too busy handling other interrupts and didn't get to that one before the next character had assived on the serial line. It got worse of course as line sppeds were increased. That's how a PDP-8/I I worked with in 1968 worked.?? I think that kind of issue is the one Alex referred to about selecting the 316 because of its interrupt mechanism.?? Since the IMP was essentially a multi-port I/O handler, how the hardware handled I/O on all those interfaces was a crucial factor in selecting the 516.?? That's why I suggested that the I/O capabilities of a microprocessor needed to be considered when trying to figure out how it compared to the 516, more so than just classic metrics like raw memory and CPU speed. About ten years ago I dug into an early version of the IMP code to figure out how it worked.?? The main CPU was essentially an interrupt processor, waiting in an idle loop for an interrupt to occur and then handling it fast enough to avoid not getting to the next interrupt fast enough to avoid losing any data. As machines matured and costs dropped, hardware interfaces became smarter and could process large chunks of data for each interrupt. Essentially the interface contained a "co-processor" that offloaded the main CPU.? I don't recall how the earliest micros handled interrupts and I/O, but that's why it's important to look at the I/O capabilities for Steve's question. /Jack Haverty On 11/15/21 1:53 PM, Noel Chiappa via Internet-history wrote: > > From: Jack Haverty > > > IIRC, another of the important criteria for selecting the Honeywell 516 > > was the I/O performance characteristics > > ... > > So in looking for the earliest "comparable" microprocessor, in addition > > to comparing metrics such as CPU speed and memory, I think you have to > > look at I/O characteristics > > Yes, but... in a _router_, the CPU didn't need to _ever_ look at most of the > data in the packet. In anything that did TCP, yeah, you had to do the > checksum, and that almost always needed the CPU to fondle each byte, but in > e.g. the CGW, since it _never_ copied packets around (not even e.g. to make > room for a longet physical network header on the front), if a packet came in > one DMA interface, and out another, the CPU never even saw most of thwe bytes > in the packet, so the CPU speed was not too relevant, it was all bus-bandwidth > dependant. > > Early on, not all network interfaces were DMA; there were three different > approaches: > > - DMA > - full packet buffers in the interface, but the CPU had to manually move > bytes from the interface to buffers in memory, so 3 bus cycles/word (with > unrolled loop with pointers to device and buffer in registers i) > instruction fetch, ii) read from interface, iii) write to memory) so 3 > times as much bus traffic per word, compared to DMA > - interrupt per word > > But even the latter wasn't _necessarily_ a problem; at MIT, the main ARPANET > gateway for quite a while used the Stanford/SRI 1822 Interface: > > https://gunkies.org/wiki/Stanford_1822_Interface > > which was interrupt/byte, but peformance wasn't a problem (that I recall). > > Like I said, for all early routers, performance was not an issue. Performance > only became an issue when there were multiple COTS router vendors, when > performance became an easy way for their marketing people to distinguish their > products from those of competitors. I doubt the users could ever have told. > > > I don't know if _early_ microprocessors (i.e. long before Motorola 68Ks, Intel > x86's, etc) supported DMA on theit memory busses; even for those that didn't, > it might have been possible to build an external bus controller, and either > stall the CPU (if the memory bus was busy doing DMA), or build a multi-port > main memory, or something. > > Noel From ajs at crankycanuck.ca Mon Nov 15 14:58:11 2021 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Mon, 15 Nov 2021 17:58:11 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> Message-ID: <20211115225811.7jn3jtzzbldh2e2w@crankycanuck.ca> Hi, On Mon, Nov 15, 2021 at 02:34:22PM -0800, Dave Crocker via Internet-history wrote: >archived under museum-quality conditions, with easy public access. Easy public access we have. Museum-quality conditions are not something, I think, that the Internet Society can guarantee at least at present. Best regards, Andrew (I work for ISOC and am speaking for it at the moment, but it was more conveneient for me to post from this address). -- Andrew Sullivan ajs at crankycanuck.ca From tte at cs.fau.de Mon Nov 15 16:23:09 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Tue, 16 Nov 2021 01:23:09 +0100 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: On Mon, Nov 15, 2021 at 11:00:53AM -0500, Clem Cole via Internet-history wrote: > Maybe, and part of my gut reaction is to agree with you, and then I think > about what Atari, Nintendo and the like did with the 6502. Someone like > Dave Hayne would be a better judge I suspect. He and his peeps did some > really amazing stuff with those early processors in those days. https://www.youtube.com/watch?v=BaTjwo1ywcI Was mandatory viewing in 1995 when it came out, amongst Amiga fans. Some interesting tidbids. Primarily that its really nice if you have chip manufacturing in-house and very short turnarounds. I don't have any numbers handy, but i started using Sun workstation from the mid-80th on and comparing to "home-computers", and i would say that during the 80th already the CPU did not matter as much as DMA based I/O as the core performance factor to 'CPU based routers'. Of course, one can see the history of packet/dma based i/o optimization almost into the 2000ths, with designs like MPLS where all you could do was really to move two DMA pointers around (start/length), and i would claim even our current protocol designs have not catched up with particle-based I/O which did start to become available in the 90th more widely (inluding even early PCI if i am not mistaken).. And late 70th/80th minicomputer designs of course predated better I/O over microcomputer designs too. Yada yada: just comparing CPU platforms won't tell us what the core historic improvement points where. > So I think we all agree, but the time of the full 16-processors like the > 8086/Z8000/68000, running on semiconductor memory, and IMP like replacement > was quite feasible. But having looked at some of the IMP code that has > leaked out, and again the existence proof of the game controllers made > using a 6502, I have a general belief that if someone has wanted to do it > and *had the right motivation*, folks like Dave and co, would have found a > way. To stick to my cunter-priority: What where the first device designs that could multiplex reading from multiple network interface effectively. Early designs AFAIK where all built on serial-port chips, and those didn't have packet buffers but just some sized byte buffers, so the CPU had to had interrupt multiplexing or round-robing polling, which limited performance. Not sure thre was any better design before the first ethnert controllers where CPU then actively could read whole packets at a time. and then DMA if i get the order right. Cheers Toerless > That motivation is still also an open question. The game control folks > were driven to keep costs as low as possible, while offering > acceptable/good performance. Something like an IMP and later a router, > less so. Performance, I would have expected to be the high order bit and > cost would have been important, but people would pay for it. Hey, the > early AGC's from Cisco were many, many thousands. > ? > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history -- --- tte at cs.fau.de From tte at cs.fau.de Mon Nov 15 16:27:58 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Tue, 16 Nov 2021 01:27:58 +0100 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <4334078A-EDEA-4219-8DAC-837DA77645F5@tony.li> References: <4334078A-EDEA-4219-8DAC-837DA77645F5@tony.li> Message-ID: But only when you finally would write that missing "IP over BGP" RFC would we be able to measure todays CPU forwarding speed for IP packets ;-)) Next deadline: April 1st 2022 ;-) On Mon, Nov 15, 2021 at 07:48:20AM -0800, Tony Li via Internet-history wrote: > Steve, > > > On Nov 15, 2021, at 6:19 AM, Steve Crocker via Internet-history wrote: > > > > Alternatively, the processing needs for a router weren't staying static, so > > perhaps matching the power of a H316 would not have been sufficient. In > > that case, the question is when did a microprocessor appear that was > > powerful enough to serve as a router? > > > We?re still waiting for that day. > > No matter what processor (or the amount of memory) that we can convince HW designers to include, it > quickly proves inadequate. Our willingness to put anything and everything into BGP makes for insatiable > demand. > > Regards, > Tony > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From brian.e.carpenter at gmail.com Mon Nov 15 17:11:38 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Tue, 16 Nov 2021 14:11:38 +1300 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> Message-ID: <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> As late as 1973, I believe you still had to go up the Digital range as far as a PDP-11/45 to get multiple levels of interrupt priority, so that a high priority interrupt could interrupt processing of a Teletype interrupt, for example. Honeywell was well ahead of that game. (As was an IBM 1800, but that had ten times the footpriint.) I never touched a 4004, but as far as I can see it only had CPU- controlled 4-bit I/O ports, with neither interrupts nor DMA. It would have been a busy little bee trying to do an IMP's job. Regards Brian On 16-Nov-21 11:54, Jack Haverty via Internet-history wrote: > True for more modern systems, but in the era of the 316/516, inexpensive > computers sometimes did I/O in the simplest possible manner.? E.g., to > handle I/O on a serial interface, the CPU might have to take an > interrupt on every byte, read that byte from the hardware interface, and > re-enable the interface quickly enough for it to be ready to handle the > next byte,?? The hardware only buffered a single byte at a time.? The > CPU also had to do that all fast enough to handle all the streams of > interrupts from all its interfaces in order not to lose data. ? That > would occur if a particular line received a byte and raised its > interrupt, but the processor was too busy handling other interrupts and > didn't get to that one before the next character had assived on the > serial line. It got worse of course as line sppeds were increased. > > That's how a PDP-8/I I worked with in 1968 worked.?? I think that kind > of issue is the one Alex referred to about selecting the 316 because of > its interrupt mechanism.?? Since the IMP was essentially a multi-port > I/O handler, how the hardware handled I/O on all those interfaces was a > crucial factor in selecting the 516.?? That's why I suggested that the > I/O capabilities of a microprocessor needed to be considered when trying > to figure out how it compared to the 516, more so than just classic > metrics like raw memory and CPU speed. About ten years ago I dug into an > early version of the IMP code to figure out how it worked.?? The main > CPU was essentially an interrupt processor, waiting in an idle loop for > an interrupt to occur and then handling it fast enough to avoid not > getting to the next interrupt fast enough to avoid losing any data. > > As machines matured and costs dropped, hardware interfaces became > smarter and could process large chunks of data for each interrupt. > Essentially the interface contained a "co-processor" that offloaded the > main CPU.? I don't recall how the earliest micros handled interrupts and > I/O, but that's why it's important to look at the I/O capabilities for > Steve's question. > > /Jack Haverty > > > On 11/15/21 1:53 PM, Noel Chiappa via Internet-history wrote: >> > From: Jack Haverty >> >> > IIRC, another of the important criteria for selecting the Honeywell 516 >> > was the I/O performance characteristics >> > ... >> > So in looking for the earliest "comparable" microprocessor, in addition >> > to comparing metrics such as CPU speed and memory, I think you have to >> > look at I/O characteristics >> >> Yes, but... in a _router_, the CPU didn't need to _ever_ look at most of the >> data in the packet. In anything that did TCP, yeah, you had to do the >> checksum, and that almost always needed the CPU to fondle each byte, but in >> e.g. the CGW, since it _never_ copied packets around (not even e.g. to make >> room for a longet physical network header on the front), if a packet came in >> one DMA interface, and out another, the CPU never even saw most of thwe bytes >> in the packet, so the CPU speed was not too relevant, it was all bus-bandwidth >> dependant. >> >> Early on, not all network interfaces were DMA; there were three different >> approaches: >> >> - DMA >> - full packet buffers in the interface, but the CPU had to manually move >> bytes from the interface to buffers in memory, so 3 bus cycles/word (with >> unrolled loop with pointers to device and buffer in registers i) >> instruction fetch, ii) read from interface, iii) write to memory) so 3 >> times as much bus traffic per word, compared to DMA >> - interrupt per word >> >> But even the latter wasn't _necessarily_ a problem; at MIT, the main ARPANET >> gateway for quite a while used the Stanford/SRI 1822 Interface: >> >> https://gunkies.org/wiki/Stanford_1822_Interface >> >> which was interrupt/byte, but peformance wasn't a problem (that I recall). >> >> Like I said, for all early routers, performance was not an issue. Performance >> only became an issue when there were multiple COTS router vendors, when >> performance became an easy way for their marketing people to distinguish their >> products from those of competitors. I doubt the users could ever have told. >> >> >> I don't know if _early_ microprocessors (i.e. long before Motorola 68Ks, Intel >> x86's, etc) supported DMA on theit memory busses; even for those that didn't, >> it might have been possible to build an external bus controller, and either >> stall the CPU (if the memory bus was busy doing DMA), or build a multi-port >> main memory, or something. >> >> Noel > > From jeanjour at comcast.net Mon Nov 15 17:19:18 2021 From: jeanjour at comcast.net (John Day) Date: Mon, 15 Nov 2021 20:19:18 -0500 Subject: [ih] Intel 4004 vs twahe IMP In-Reply-To: <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> Message-ID: That was true on the early PDP-11 (later called the PDP-11/20), i.e. an interrupt for each character. But within a year or two, they had a DMA for serial lines. John > On Nov 15, 2021, at 17:54, Jack Haverty via Internet-history wrote: > > True for more modern systems, but in the era of the 316/516, inexpensive computers sometimes did I/O in the simplest possible manner. E.g., to handle I/O on a serial interface, the CPU might have to take an interrupt on every byte, read that byte from the hardware interface, and re-enable the interface quickly enough for it to be ready to handle the next byte, The hardware only buffered a single byte at a time. The CPU also had to do that all fast enough to handle all the streams of interrupts from all its interfaces in order not to lose data. That would occur if a particular line received a byte and raised its interrupt, but the processor was too busy handling other interrupts and didn't get to that one before the next character had assived on the serial line. It got worse of course as line sppeds were increased. > > That's how a PDP-8/I I worked with in 1968 worked. I think that kind of issue is the one Alex referred to about selecting the 316 because of its interrupt mechanism. Since the IMP was essentially a multi-port I/O handler, how the hardware handled I/O on all those interfaces was a crucial factor in selecting the 516. That's why I suggested that the I/O capabilities of a microprocessor needed to be considered when trying to figure out how it compared to the 516, more so than just classic metrics like raw memory and CPU speed. About ten years ago I dug into an early version of the IMP code to figure out how it worked. The main CPU was essentially an interrupt processor, waiting in an idle loop for an interrupt to occur and then handling it fast enough to avoid not getting to the next interrupt fast enough to avoid losing any data. > > As machines matured and costs dropped, hardware interfaces became smarter and could process large chunks of data for each interrupt. Essentially the interface contained a "co-processor" that offloaded the main CPU. I don't recall how the earliest micros handled interrupts and I/O, but that's why it's important to look at the I/O capabilities for Steve's question. > > /Jack Haverty > > > On 11/15/21 1:53 PM, Noel Chiappa via Internet-history wrote: >> > From: Jack Haverty >> >> > IIRC, another of the important criteria for selecting the Honeywell 516 >> > was the I/O performance characteristics >> > ... >> > So in looking for the earliest "comparable" microprocessor, in addition >> > to comparing metrics such as CPU speed and memory, I think you have to >> > look at I/O characteristics >> >> Yes, but... in a _router_, the CPU didn't need to _ever_ look at most of the >> data in the packet. In anything that did TCP, yeah, you had to do the >> checksum, and that almost always needed the CPU to fondle each byte, but in >> e.g. the CGW, since it _never_ copied packets around (not even e.g. to make >> room for a longet physical network header on the front), if a packet came in >> one DMA interface, and out another, the CPU never even saw most of thwe bytes >> in the packet, so the CPU speed was not too relevant, it was all bus-bandwidth >> dependant. >> >> Early on, not all network interfaces were DMA; there were three different >> approaches: >> >> - DMA >> - full packet buffers in the interface, but the CPU had to manually move >> bytes from the interface to buffers in memory, so 3 bus cycles/word (with >> unrolled loop with pointers to device and buffer in registers i) >> instruction fetch, ii) read from interface, iii) write to memory) so 3 >> times as much bus traffic per word, compared to DMA >> - interrupt per word >> >> But even the latter wasn't _necessarily_ a problem; at MIT, the main ARPANET >> gateway for quite a while used the Stanford/SRI 1822 Interface: >> >> https://gunkies.org/wiki/Stanford_1822_Interface >> >> which was interrupt/byte, but peformance wasn't a problem (that I recall). >> >> Like I said, for all early routers, performance was not an issue. Performance >> only became an issue when there were multiple COTS router vendors, when >> performance became an easy way for their marketing people to distinguish their >> products from those of competitors. I doubt the users could ever have told. >> >> >> I don't know if _early_ microprocessors (i.e. long before Motorola 68Ks, Intel >> x86's, etc) supported DMA on theit memory busses; even for those that didn't, >> it might have been possible to build an external bus controller, and either >> stall the CPU (if the memory bus was busy doing DMA), or build a multi-port >> main memory, or something. >> >> Noel > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From salo at saloits.com Mon Nov 15 17:35:31 2021 From: salo at saloits.com (Timothy J. Salo) Date: Mon, 15 Nov 2021 19:35:31 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> Message-ID: This discussion parallels discussions I had in the 1980s at my then-employer, which build communications processors for mainframe computers. Microprocessors were proliferating and quickly gaining in power. As has been noted, microprocessors typically lacked the I/O capabilities, throughput and interrupt structure, of purpose-built communications processors. But, the real problem with microprocessors, for those who had existing communications processors and software, was that the microprocessors had instruction sets that were incompatible with existing communications processors. No mater what the benefits of microprocessors, we had millions of lines of assembly code that would have been pretty much impossible to port to a microprocessor. We never tried to port our software to a 68000 or any other processor. I believe that BBN was in a similar position with the IMP software, although obviously there are people here who are far more aware of this than am I. If I recall correctly, and my information was pretty indirect, BBN was trying to port their IMP software to a 68000. Was this the Butterfly? But, processors such as the 68000s provided an excellent platform for companies that weren't burdened by millions of lines of assembly language software. Most notably Cisco. And, we all know how that story turned out. -tjs From dhc at dcrocker.net Mon Nov 15 17:37:36 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Mon, 15 Nov 2021 17:37:36 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <20211115225811.7jn3jtzzbldh2e2w@crankycanuck.ca> References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> <20211115225811.7jn3jtzzbldh2e2w@crankycanuck.ca> Message-ID: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> On 11/15/2021 2:58 PM, Andrew Sullivan via Internet-history wrote: > Easy public access we have.? Museum-quality conditions are not > something, I think, that the Internet Society can guarantee at least at > present. I must admit that it had not occurred to me that ISOC might be the provider of museum-quality archiving and access. I'd rather assumed it might be done by a... museum. d/ ps. I got a private note asking about a paid archive at archive.org. I think that a reasonable question and possibly worth doing on it own. Whether that qualifies as 'museum quality' isn't something I can assess. -- Dave Crocker Brandenburg InternetWorking bbiw.net From jmamodio at gmail.com Mon Nov 15 17:47:25 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 19:47:25 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> Message-ID: The 4004 was very limited, it was a 16 pin chip, 4-bit share bus, able to address only 32K of ROM and 5,120 bits of RAM, yes bits, not bytes. ALU optimized for BCD arithmetic, after all was designed for a calculator. Regards Jorge On Mon, Nov 15, 2021 at 7:11 PM Brian E Carpenter via Internet-history < internet-history at elists.isoc.org> wrote: > As late as 1973, I believe you still had to go up the Digital > range as far as a PDP-11/45 to get multiple levels of interrupt > priority, so that a high priority interrupt could interrupt > processing of a Teletype interrupt, for example. Honeywell > was well ahead of that game. (As was an IBM 1800, but that had > ten times the footpriint.) > > I never touched a 4004, but as far as I can see it only had CPU- > controlled 4-bit I/O ports, with neither interrupts nor DMA. It > would have been a busy little bee trying to do an IMP's job. > > Regards > Brian > > On 16-Nov-21 11:54, Jack Haverty via Internet-history wrote: > > True for more modern systems, but in the era of the 316/516, inexpensive > > computers sometimes did I/O in the simplest possible manner. E.g., to > > handle I/O on a serial interface, the CPU might have to take an > > interrupt on every byte, read that byte from the hardware interface, and > > re-enable the interface quickly enough for it to be ready to handle the > > next byte, The hardware only buffered a single byte at a time. The > > CPU also had to do that all fast enough to handle all the streams of > > interrupts from all its interfaces in order not to lose data. That > > would occur if a particular line received a byte and raised its > > interrupt, but the processor was too busy handling other interrupts and > > didn't get to that one before the next character had assived on the > > serial line. It got worse of course as line sppeds were increased. > > > > That's how a PDP-8/I I worked with in 1968 worked. I think that kind > > of issue is the one Alex referred to about selecting the 316 because of > > its interrupt mechanism. Since the IMP was essentially a multi-port > > I/O handler, how the hardware handled I/O on all those interfaces was a > > crucial factor in selecting the 516. That's why I suggested > that the > > I/O capabilities of a microprocessor needed to be considered when trying > > to figure out how it compared to the 516, more so than just classic > > metrics like raw memory and CPU speed. About ten years ago I dug into an > > early version of the IMP code to figure out how it worked. The main > > CPU was essentially an interrupt processor, waiting in an idle loop for > > an interrupt to occur and then handling it fast enough to avoid not > > getting to the next interrupt fast enough to avoid losing any data. > > > > As machines matured and costs dropped, hardware interfaces became > > smarter and could process large chunks of data for each interrupt. > > Essentially the interface contained a "co-processor" that offloaded the > > main CPU. I don't recall how the earliest micros handled interrupts and > > I/O, but that's why it's important to look at the I/O capabilities for > > Steve's question. > > > > /Jack Haverty > > > > > > On 11/15/21 1:53 PM, Noel Chiappa via Internet-history wrote: > >> > From: Jack Haverty > >> > >> > IIRC, another of the important criteria for selecting the > Honeywell 516 > >> > was the I/O performance characteristics > >> > ... > >> > So in looking for the earliest "comparable" microprocessor, in > addition > >> > to comparing metrics such as CPU speed and memory, I think you > have to > >> > look at I/O characteristics > >> > >> Yes, but... in a _router_, the CPU didn't need to _ever_ look at most > of the > >> data in the packet. In anything that did TCP, yeah, you had to do the > >> checksum, and that almost always needed the CPU to fondle each byte, > but in > >> e.g. the CGW, since it _never_ copied packets around (not even e.g. to > make > >> room for a longet physical network header on the front), if a packet > came in > >> one DMA interface, and out another, the CPU never even saw most of thwe > bytes > >> in the packet, so the CPU speed was not too relevant, it was all > bus-bandwidth > >> dependant. > >> > >> Early on, not all network interfaces were DMA; there were three > different > >> approaches: > >> > >> - DMA > >> - full packet buffers in the interface, but the CPU had to manually > move > >> bytes from the interface to buffers in memory, so 3 bus > cycles/word (with > >> unrolled loop with pointers to device and buffer in registers i) > >> instruction fetch, ii) read from interface, iii) write to memory) > so 3 > >> times as much bus traffic per word, compared to DMA > >> - interrupt per word > >> > >> But even the latter wasn't _necessarily_ a problem; at MIT, the main > ARPANET > >> gateway for quite a while used the Stanford/SRI 1822 Interface: > >> > >> https://gunkies.org/wiki/Stanford_1822_Interface > >> > >> which was interrupt/byte, but peformance wasn't a problem (that I > recall). > >> > >> Like I said, for all early routers, performance was not an issue. > Performance > >> only became an issue when there were multiple COTS router vendors, when > >> performance became an easy way for their marketing people to > distinguish their > >> products from those of competitors. I doubt the users could ever have > told. > >> > >> > >> I don't know if _early_ microprocessors (i.e. long before Motorola > 68Ks, Intel > >> x86's, etc) supported DMA on theit memory busses; even for those that > didn't, > >> it might have been possible to build an external bus controller, and > either > >> stall the CPU (if the memory bus was busy doing DMA), or build a > multi-port > >> main memory, or something. > >> > >> Noel > > > > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From salo at saloits.com Mon Nov 15 17:48:22 2021 From: salo at saloits.com (Timothy J. Salo) Date: Mon, 15 Nov 2021 19:48:22 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: <396bc3ed-1602-5448-530e-5c0214ea13ea@saloits.com> On 11/15/2021 10:00 AM, Clem Cole via Internet-history wrote: > So I think we all agree, but the time of the full 16-processors like the > 8086/Z8000/68000, running on semiconductor memory, and IMP like replacement > was quite feasible. But having looked at some of the IMP code that has > leaked out, and again the existence proof of the game controllers made > using a 6502, I have a general belief that if someone has wanted to do it > and *had the right motivation*, folks like Dave and co, would have found a > way. In my view, that is exactly what Cisco did. But, they had the substantial advantage of not being burdened by an existing code base. -tjs From salo at saloits.com Mon Nov 15 18:11:13 2021 From: salo at saloits.com (Timothy J. Salo) Date: Mon, 15 Nov 2021 20:11:13 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <396bc3ed-1602-5448-530e-5c0214ea13ea@saloits.com> References: <396bc3ed-1602-5448-530e-5c0214ea13ea@saloits.com> Message-ID: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> I think we all agree that the IMP was a pretty limited machine. From the backup slides of a presentation of mine: Early ARPANET router, Interface Message Processor (IMP), (1969): o 16-bit words, 12-16 K-word memory o 100-?sec clock (10 KHz) o Early ARPANET links: 56 kbps o 0.18 clock cycles per bit I have argued that this, 12-16 K words of memory, is why we had the end-to-end argument (which morphed into a principal and then into a canon). (The rest of the presentation pretty much ignores the end-to-end argument.) Also from this presentation: Early NSFNET router: DEC LSI-11/73 (1983) with Fuzzball router o 512 KB memory o (15.2 MHz) o 271 clock cycles/bit -tjs From ajs at crankycanuck.ca Mon Nov 15 18:33:22 2021 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Mon, 15 Nov 2021 21:33:22 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> Message-ID: <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> Ah, ok. Well, the archive we are operating (or, really, paying someone else to archive) is what we are able to offer right now. We do archive the Internet Society pages at archive.org. I am not sure whether that extends to list archives but I will find out. A ? Andrew Sullivan Please excuse my clumbsy thums > On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history wrote: > > ?On 11/15/2021 2:58 PM, Andrew Sullivan via Internet-history wrote: >> Easy public access we have. Museum-quality conditions are not something, I think, that the Internet Society can guarantee at least at present. > > > I must admit that it had not occurred to me that ISOC might be the provider of museum-quality archiving and access. > > I'd rather assumed it might be done by a... museum. > > d/ > > ps. I got a private note asking about a paid archive at archive.org. I think that a reasonable question and possibly worth doing on it own. Whether that qualifies as 'museum quality' isn't something I can assess. > > > -- > Dave Crocker > Brandenburg InternetWorking > bbiw.net > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From steve at shinkuro.com Mon Nov 15 18:35:02 2021 From: steve at shinkuro.com (Steve Crocker) Date: Mon, 15 Nov 2021 21:35:02 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> References: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> Message-ID: <0271C632-96CF-4605-B9FD-9F43040687E1@shinkuro.com> Arpanet lines were 50 kbs, not 56 kbs. Sent from my iPhone > On Nov 15, 2021, at 9:11 PM, Timothy J. Salo via Internet-history wrote: > > ?I think we all agree that the IMP was a pretty limited machine. From > the backup slides of a presentation of mine: > > Early ARPANET router, Interface Message Processor (IMP), (1969): > o 16-bit words, 12-16 K-word memory > o 100-?sec clock (10 KHz) > o Early ARPANET links: 56 kbps > o 0.18 clock cycles per bit > > I have argued that this, 12-16 K words of memory, is why we had the > end-to-end argument (which morphed into a principal and then into a > canon). > > (The rest of the presentation pretty much ignores the end-to-end > argument.) > > Also from this presentation: > > Early NSFNET router: DEC LSI-11/73 (1983) with Fuzzball router > o 512 KB memory > o (15.2 MHz) > o 271 clock cycles/bit > > -tjs > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From woody at pch.net Mon Nov 15 18:53:13 2021 From: woody at pch.net (Bill Woodcock) Date: Tue, 16 Nov 2021 03:53:13 +0100 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> Message-ID: >> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history wrote: >> I got a private note asking about a paid archive at archive.org. I think that a reasonable question and possibly worth doing on it own. Whether that qualifies as 'museum quality' isn't something I can assess. Well, it?s what we?ve got for ?archival quality? on the Internet right now, and if it?s not good enough, that means Brewster needs more support. -Bill -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 833 bytes Desc: Message signed with OpenPGP URL: From brian.e.carpenter at gmail.com Mon Nov 15 18:54:07 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Tue, 16 Nov 2021 15:54:07 +1300 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> Message-ID: I told a lie. The PDP-11/20 did have priority interrupts, and the Unibus was intrinsically capable of DMA. So it probably could have done the IMP job, had it been on the market a year or three earlier. But you needed an 11/45 to do any real multiprogramming with memory management. Regards Brian On 16-Nov-21 14:11, Brian E Carpenter wrote: > As late as 1973, I believe you still had to go up the Digital > range as far as a PDP-11/45 to get multiple levels of interrupt > priority, so that a high priority interrupt could interrupt > processing of a Teletype interrupt, for example. Honeywell > was well ahead of that game. (As was an IBM 1800, but that had > ten times the footpriint.) > > I never touched a 4004, but as far as I can see it only had CPU- > controlled 4-bit I/O ports, with neither interrupts nor DMA. It > would have been a busy little bee trying to do an IMP's job. > > Regards > Brian > > On 16-Nov-21 11:54, Jack Haverty via Internet-history wrote: >> True for more modern systems, but in the era of the 316/516, inexpensive >> computers sometimes did I/O in the simplest possible manner.? E.g., to >> handle I/O on a serial interface, the CPU might have to take an >> interrupt on every byte, read that byte from the hardware interface, and >> re-enable the interface quickly enough for it to be ready to handle the >> next byte,?? The hardware only buffered a single byte at a time.? The >> CPU also had to do that all fast enough to handle all the streams of >> interrupts from all its interfaces in order not to lose data. ? That >> would occur if a particular line received a byte and raised its >> interrupt, but the processor was too busy handling other interrupts and >> didn't get to that one before the next character had assived on the >> serial line. It got worse of course as line sppeds were increased. >> >> That's how a PDP-8/I I worked with in 1968 worked.?? I think that kind >> of issue is the one Alex referred to about selecting the 316 because of >> its interrupt mechanism.?? Since the IMP was essentially a multi-port >> I/O handler, how the hardware handled I/O on all those interfaces was a >> crucial factor in selecting the 516.?? That's why I suggested > that the >> I/O capabilities of a microprocessor needed to be considered when trying >> to figure out how it compared to the 516, more so than just classic >> metrics like raw memory and CPU speed. About ten years ago I dug into an >> early version of the IMP code to figure out how it worked.?? The main >> CPU was essentially an interrupt processor, waiting in an idle loop for >> an interrupt to occur and then handling it fast enough to avoid not >> getting to the next interrupt fast enough to avoid losing any data. >> >> As machines matured and costs dropped, hardware interfaces became >> smarter and could process large chunks of data for each interrupt. >> Essentially the interface contained a "co-processor" that offloaded the >> main CPU.? I don't recall how the earliest micros handled interrupts and >> I/O, but that's why it's important to look at the I/O capabilities for >> Steve's question. >> >> /Jack Haverty >> >> >> On 11/15/21 1:53 PM, Noel Chiappa via Internet-history wrote: >>> > From: Jack Haverty >>> >>> > IIRC, another of the important criteria for selecting the Honeywell 516 >>> > was the I/O performance characteristics >>> > ... >>> > So in looking for the earliest "comparable" microprocessor, in > addition >>> > to comparing metrics such as CPU speed and memory, I think you > have to >>> > look at I/O characteristics >>> >>> Yes, but... in a _router_, the CPU didn't need to _ever_ look at most of the >>> data in the packet. In anything that did TCP, yeah, you had to do the >>> checksum, and that almost always needed the CPU to fondle each byte, but in >>> e.g. the CGW, since it _never_ copied packets around (not even e.g. to > make >>> room for a longet physical network header on the front), if a packet came in >>> one DMA interface, and out another, the CPU never even saw most of thwe bytes >>> in the packet, so the CPU speed was not too relevant, it was all bus-bandwidth >>> dependant. >>> >>> Early on, not all network interfaces were DMA; there were three different >>> approaches: >>> >>> - DMA >>> - full packet buffers in the interface, but the CPU had to manually > move >>> bytes from the interface to buffers in memory, so 3 bus cycles/word (with >>> unrolled loop with pointers to device and buffer in registers i) >>> instruction fetch, ii) read from interface, iii) write to memory) > so 3 >>> times as much bus traffic per word, compared to DMA >>> - interrupt per word >>> >>> But even the latter wasn't _necessarily_ a problem; at MIT, the main ARPANET >>> gateway for quite a while used the Stanford/SRI 1822 Interface: >>> >>> https://gunkies.org/wiki/Stanford_1822_Interface >>> >>> which was interrupt/byte, but peformance wasn't a problem (that I recall). >>> >>> Like I said, for all early routers, performance was not an issue. Performance >>> only became an issue when there were multiple COTS router vendors, when >>> performance became an easy way for their marketing people to distinguish their >>> products from those of competitors. I doubt the users could ever have told. >>> >>> >>> I don't know if _early_ microprocessors (i.e. long before Motorola 68Ks, Intel >>> x86's, etc) supported DMA on theit memory busses; even for those that didn't, >>> it might have been possible to build an external bus controller, and either >>> stall the CPU (if the memory bus was busy doing DMA), or build a multi-port >>> main memory, or something. >>> >>> Noel >> >> > From touch at strayalpha.com Mon Nov 15 18:57:13 2021 From: touch at strayalpha.com (touch at strayalpha.com) Date: Mon, 15 Nov 2021 18:57:13 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> Message-ID: Hi, Dave (et al.), Your friendly neighborhood list originator and admin here. > On Nov 15, 2021, at 2:34 PM, Dave Crocker via Internet-history wrote: > > I don't remember whether this has been discussed or resolved here, before, but looking at the depth of historical knowledge that is casually proffered here makes me wish that the list is automatically archived under museum-quality conditions, with easy public access. It has been several times, notably when we moved it from the unstable resources at ISI (that did not even include off-site backup) to the ISOC. I have tried many times to find a more archival home for the list, but: - we do not monetize info about list posters or content in any way - we do not charge to participate in the list Thus we have no funds to pay for such a service, were one to exist. Note that archive.org does not run mailing lists AND its services require a per-year fee. They don?t say what happens when you stop paying, but in-perpituity escrows aren?t cheap to setup, even with initial funds. HOWEVER, note that such an service has other constraints, notably: - honors the ?no monetization? conditions above - does not limit posts per day or number of subscribers - allows full control by the list admin - can support a transfer of the current list archives I?ve contacted many more typical archive-friendly places, including the Computer History Museum, but nobody currently runs a list intended to be archival, so none were able to assist. Such lists hosted at the IEEE ComSoc or ACM SIGCOMM societies have no archive capability either. The closest such ?archive? is that this list, being on the Internet, may be incidentally archived by places like the Internet Wayback machine. Joe From jnc at mercury.lcs.mit.edu Mon Nov 15 18:58:48 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Mon, 15 Nov 2021 21:58:48 -0500 (EST) Subject: [ih] Intel 4004 vs the IMP Message-ID: <20211116025848.E3A4D18C090@mercury.lcs.mit.edu> > E.g. a register-register MOV took 1.7 microseconds, but a memory-memory > move .. seems to have taken 1.7+4.9=6.6 usec, so about 1/5th of a MIP Ooops, I mistakenly was in the -11/23 timing appendix, not the -11/03; the latter was considerably slower: 3.5 usec basic, 2.5(sic)+9.1=11.6 usec. (Indirect was even slower.) So about 1/10th of a MIP. > From: Jack Haverty > E.g., to handle I/O on a serial interface, the CPU might have to take > an interrupt on every byte > ... > I think that kind of issue is the one Alex referred to about selecting > the 316 because of its interrupt mechanism. ... how the hardware > handled I/O on all those interfaces was a crucial factor in selecting > the 516. Yes and no. The IMP's modem and host interfaces were both DMA (in the sense that the CPU only got a single interrupt - which diverted instruction processing in the CPU to other instructions - for every packet, not on every word; the details differed significantly from modern DMA, though - see below). But there was a timing issue. Per 'The interface message processor for the ARPA computer network', available here: https://www.walden-family.com/public/1970-imp-afips.pdf "To send a packet, the IMP program sets up memory pointers to the packet and then activates the interface ... The interface takes successive words from the memory using its assigned output data channel and transmits them bit-serially (to the Host or to the modem). When the memory buffer has thus been emptied, the interface notifies the program via an interrupt". The details are intetesting: the IMP used "a set of 16 multiplexed channels (which implement a 4-cycle data break)". This is through a device called the DMC, the 'Direct Multiplex Control' (sometimes 'Data Multiplex Control'). Notice the "4-cycle data break"; consulting the DMC manual (also online), this was very similar to the '3-cycle data break' used on many early DEC machines, up through the PDP-8. This kept the buffer address and count in main memory (to reduce to cost of devices); the downside is that it increased the memory bandwidth usage. (The 4 cycles were 1) current buffer address read, ii) buffer extent read, iii) data read/write, iv) modified buffer address write-back.) (Hey, it could have been worse; the DM11 asynchronous line interface of the early PDP-11 kept the _shift registers_ in main memory, and used DMA to gain access to them during input/output. At least it was efficient DMA - the memory address was stored in the device! :-) The timing issue comes from the fact that, as far as I can tell from the IMP hardware manual: https://walden-family.com/impcode/imp-hardware.pdf there was _no_ buffering on the modem and host interfaces, just the shift register; not so bad on the host interface, which used a handshake, and could be paused, but potentially problematic on the synchronous modem interface; after a word arrived, it had to be written to memory before the first _bit_ of the next word arrived. (The DM11 had the same issue.) > That's why I suggested that the I/O capabilities of a microprocessor > needed to be considered when trying to figure out how it compared to > the 516, more so than just classic metrics like raw memory and CPU speed. That part I agree with. (But don't forget the address space, either; the 4004 really had too small an address space to be usable as a router at _any point in time_.) Noel From jmamodio at gmail.com Mon Nov 15 19:00:20 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 21:00:20 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <0271C632-96CF-4605-B9FD-9F43040687E1@shinkuro.com> References: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> <0271C632-96CF-4605-B9FD-9F43040687E1@shinkuro.com> Message-ID: Weren't the BBN modems capable of handling up to 6 "phone lines" at 50,000bps. So basically analog dedicated phone lines. I'm sure the early lines were not DDS 56K Regards Jorge On Mon, Nov 15, 2021 at 8:35 PM Steve Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > Arpanet lines were 50 kbs, not 56 kbs. > > Sent from my iPhone > > > On Nov 15, 2021, at 9:11 PM, Timothy J. Salo via Internet-history < > internet-history at elists.isoc.org> wrote: > > > > ?I think we all agree that the IMP was a pretty limited machine. From > > the backup slides of a presentation of mine: > > > > Early ARPANET router, Interface Message Processor (IMP), (1969): > > o 16-bit words, 12-16 K-word memory > > o 100-?sec clock (10 KHz) > > o Early ARPANET links: 56 kbps > > o 0.18 clock cycles per bit > > > > I have argued that this, 12-16 K words of memory, is why we had the > > end-to-end argument (which morphed into a principal and then into a > > canon). > > > > (The rest of the presentation pretty much ignores the end-to-end > > argument.) > > > > Also from this presentation: > > > > Early NSFNET router: DEC LSI-11/73 (1983) with Fuzzball router > > o 512 KB memory > > o (15.2 MHz) > > o 271 clock cycles/bit > > > > -tjs > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From steve at shinkuro.com Mon Nov 15 19:07:05 2021 From: steve at shinkuro.com (Steve Crocker) Date: Mon, 15 Nov 2021 22:07:05 -0500 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> <0271C632-96CF-4605-B9FD-9F43040687E1@shinkuro.com> Message-ID: I believe the first batch of IMPs were capable of handling up to four hosts and up to four 50kbs communication lines, but the total was limited to 7. The limitation of 7 interfaces was due to the physical slots inside the machine, not a bandwidth limitation. And, yes, the communication lines were 50 kbs using Western Electric 303A modems. The modem sprayed the data across 12 twelve(!) voice grade analog lines. Digital communication lines came along later. Steve On Mon, Nov 15, 2021 at 10:00 PM Jorge Amodio wrote: > > Weren't the BBN modems capable of handling up to 6 "phone lines" at > 50,000bps. So basically analog dedicated phone lines. > > I'm sure the early lines were not DDS 56K > > Regards > Jorge > > > On Mon, Nov 15, 2021 at 8:35 PM Steve Crocker via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Arpanet lines were 50 kbs, not 56 kbs. >> >> Sent from my iPhone >> >> > On Nov 15, 2021, at 9:11 PM, Timothy J. Salo via Internet-history < >> internet-history at elists.isoc.org> wrote: >> > >> > ?I think we all agree that the IMP was a pretty limited machine. From >> > the backup slides of a presentation of mine: >> > >> > Early ARPANET router, Interface Message Processor (IMP), (1969): >> > o 16-bit words, 12-16 K-word memory >> > o 100-?sec clock (10 KHz) >> > o Early ARPANET links: 56 kbps >> > o 0.18 clock cycles per bit >> > >> > I have argued that this, 12-16 K words of memory, is why we had the >> > end-to-end argument (which morphed into a principal and then into a >> > canon). >> > >> > (The rest of the presentation pretty much ignores the end-to-end >> > argument.) >> > >> > Also from this presentation: >> > >> > Early NSFNET router: DEC LSI-11/73 (1983) with Fuzzball router >> > o 512 KB memory >> > o (15.2 MHz) >> > o 271 clock cycles/bit >> > >> > -tjs >> > -- >> > Internet-history mailing list >> > Internet-history at elists.isoc.org >> > https://elists.isoc.org/mailman/listinfo/internet-history >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > From jmamodio at gmail.com Mon Nov 15 19:07:01 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 21:07:01 -0600 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> <0271C632-96CF-4605-B9FD-9F43040687E1@shinkuro.com> Message-ID: Yup, 6 modem interfaces on the IMP design, I remember that it was on one of the early reports from BBN #1763 (1969). Here you go, a very detailed description of the IMP design by BBN. https://walden-family.com/impcode/1969-initial-IMP-design.pdf Cheers Jorge On Mon, Nov 15, 2021 at 9:00 PM Jorge Amodio wrote: > > Weren't the BBN modems capable of handling up to 6 "phone lines" at > 50,000bps. So basically analog dedicated phone lines. > > I'm sure the early lines were not DDS 56K > > Regards > Jorge > > > On Mon, Nov 15, 2021 at 8:35 PM Steve Crocker via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Arpanet lines were 50 kbs, not 56 kbs. >> >> Sent from my iPhone >> >> > On Nov 15, 2021, at 9:11 PM, Timothy J. Salo via Internet-history < >> internet-history at elists.isoc.org> wrote: >> > >> > ?I think we all agree that the IMP was a pretty limited machine. From >> > the backup slides of a presentation of mine: >> > >> > Early ARPANET router, Interface Message Processor (IMP), (1969): >> > o 16-bit words, 12-16 K-word memory >> > o 100-?sec clock (10 KHz) >> > o Early ARPANET links: 56 kbps >> > o 0.18 clock cycles per bit >> > >> > I have argued that this, 12-16 K words of memory, is why we had the >> > end-to-end argument (which morphed into a principal and then into a >> > canon). >> > >> > (The rest of the presentation pretty much ignores the end-to-end >> > argument.) >> > >> > Also from this presentation: >> > >> > Early NSFNET router: DEC LSI-11/73 (1983) with Fuzzball router >> > o 512 KB memory >> > o (15.2 MHz) >> > o 271 clock cycles/bit >> > >> > -tjs >> > -- >> > Internet-history mailing list >> > Internet-history at elists.isoc.org >> > https://elists.isoc.org/mailman/listinfo/internet-history >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > From dhc at dcrocker.net Mon Nov 15 19:08:10 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Mon, 15 Nov 2021 19:08:10 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> Message-ID: On 11/15/2021 6:57 PM, touch at strayalpha.com wrote: > I?ve contacted many more typical archive-friendly places, including the > Computer History Museum, but nobody currently runs a list intended to be > archival, so none were able to assist. Joe, nice summary of issue.s Thanks. Maybe it's just me, but this seems a topic worthy of solution. Not just for this list, of course. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From jack at 3kitty.org Mon Nov 15 19:09:06 2021 From: jack at 3kitty.org (Jack Haverty) Date: Mon, 15 Nov 2021 19:09:06 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: Message-ID: <934040ee-2d99-2689-2a41-d8ccf35a18a7@3kitty.org> On 11/15/21 8:00 AM, Clem Cole via Internet-history wrote: > But having looked at some of the IMP code that has > leaked out, ?I wouldn't say "leaked out".??? It was methodically resurrected and Dave Walden (who was one of the programmers) has put it on his website.?? There is quite a lot of information about the IMP code, including several complete listings of its program, in 316 assembler of course, and its recent resurrection, available at: https://walden-family.com/impcode/ I'm not sure where it is now, but the code itself also exists in runnable form and has been successfully run, using a Honeywell 316 emulator, to recreate the original 4-node early ARPANET configuration.?? It's ready for someone to resurrect one of those early host computers to create traffic. I had the opportunity to dive pretty deeply into the code itself while working on a patent dispute about ten years ago, and figure out exactly how it did what we remember it could do.?? It's a fascinating and impressive piece of work.?? The IMP software probably violated lots of principles of modern programming practice, e.g., using techniques like self-modifying code. ? But it worked. You did what you had to do with the hardware that was available. Anyone who has the interest, curiosity, and time to look at a historic chunk of actual code, I suggest looking at that IMP code. It really should be in a museum. /Jack Haverty From internet-history at gtaylor.tnetconsulting.net Mon Nov 15 19:18:47 2021 From: internet-history at gtaylor.tnetconsulting.net (Grant Taylor) Date: Mon, 15 Nov 2021 20:18:47 -0700 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> Message-ID: <704d4a6c-7498-64c2-ce7f-b2002109ecbc@spamtrap.tnetconsulting.net> All very good points Dave. On 11/15/21 7:57 PM, touch--- via Internet-history wrote: > I?ve contacted many more typical archive-friendly places, including > the Computer History Museum, but nobody currently runs a list intended > to be archival, so none were able to assist. Such lists hosted at the > IEEE ComSoc or ACM SIGCOMM societies have no archive capability either. I wanted to point out that running the list and archiving the list are two distinct operations that could be done by independent entities. As in the list archive could be it's own entity that is subscribed to a completely independent list that is run by a completely independent entity. -- Grant. . . . unix || die From tony.li at tony.li Mon Nov 15 19:19:15 2021 From: tony.li at tony.li (Tony Li) Date: Mon, 15 Nov 2021 19:19:15 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> Message-ID: > But, processors such as the 68000s provided an excellent platform for > companies that weren't burdened by millions of lines of assembly > language software. Most notably Cisco. And, we all know how that > story turned out. If you?re implying that the 68000 somehow led to Cisco?s success, you?d be ignoring the real workhorse, the TI DSP chip that was at the heart of Cisco?s performance. This was the slave co-processor on the original MCI card and took care of moving the bytes, managing the queues, doing the IP checksum and fixup, and doing the MAC rewrite. The 68k CPU was only responsible for doing the IP address lookup. This is what allowed Cisco to get to 10Mbps Ethernet wire speed performance. This was extended into the AGS+ cBus architecture, where again the DSP did the lion?s share of the work and even took over doing the address lookup. Even the DSP was insufficient, so it was assisted by additional FPGA based processing and then (after a major wrong turn) things transitioned to an ASIC based forwarding architecture. Microprocessors for all of their benefits have never been sufficient. More silicon has always been a fundamental requirement. T From touch at strayalpha.com Mon Nov 15 19:24:54 2021 From: touch at strayalpha.com (touch at strayalpha.com) Date: Mon, 15 Nov 2021 19:24:54 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> Message-ID: <44AC7817-F588-4537-B56C-D893BE315029@strayalpha.com> ? Joe Touch, temporal epistemologist www.strayalpha.com > On Nov 15, 2021, at 7:08 PM, Dave Crocker wrote: > > On 11/15/2021 6:57 PM, touch at strayalpha.com wrote: >> I?ve contacted many more typical archive-friendly places, including the Computer History Museum, but nobody currently runs a list intended to be archival, so none were able to assist. > > Joe, nice summary of issue.s Thanks. > > Maybe it's just me, but this seems a topic worthy of solution. Not just for this list, of course. It does, but the idea of having one?s online presence ?passed on? is very recent (Apple?s ?digital legacy? program is claimed to arrive in IOS 15.2). There are a few services that offer ?permanent escrow? of data for various fees, but companies come and go - so YYMV. Here?s the thing: this doesn?t exist IRL either. Bodily internment isn?t forever; cemeteries get bought and sold, sometimes even converted into other types of property (malls, homes, e.g. - Poltergeist). Same for museums - they come and go, collections are bought and sold and some disappear forever. And remember - we?re not looking for a static archive; we?re really looking for a place we can operate AND be archival. They may be separate solutions, but have to work together. Joe From johnl at iecc.com Mon Nov 15 19:38:47 2021 From: johnl at iecc.com (John Levine) Date: 15 Nov 2021 22:38:47 -0500 Subject: [ih] Intel 4004 vs the IMP and other ancient history In-Reply-To: <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> Message-ID: <20211116033849.2D0742FFFC3E@ary.qy> It appears that Brian E Carpenter via Internet-history said: >As late as 1973, I believe you still had to go up the Digital >range as far as a PDP-11/45 to get multiple levels of interrupt No, the original 11/20 had a four-level priority interrupt, numbered BR4 to BR7 plus a higher priority NPR (non-processor request) to do DMA. The 11/45 added levels BR1 through 3. >On 16-Nov-21 11:54, Jack Haverty via Internet-history wrote: >> True for more modern systems, but in the era of the 316/516, inexpensive >> computers sometimes did I/O in the simplest possible manner.? E.g., to >> handle I/O on a serial interface, the CPU might have to take an >> interrupt on every byte, read that byte from the hardware interface, and >> re-enable the interface quickly enough for it to be ready to handle the >> next byte,?... >> >> That's how a PDP-8/I I worked with in 1968 worked.?... Depends on how much you wanted to pay. The PDP-8 and 8/I had single cycle data break where the device provided the data address and transferred a word in or out of memory, and the slower but cheaper three-cycle data break where the device provided a fixed address (programmed with diodes I think) of two words which were the word count and data address that the CPU updated, did the transfer, and then told the device if the count incremented to zero. Logic was expensive in those days so if you could get by with an interrupt per character or word you did. The disk, drum, and DECtape controllers used data break, but the 680 multiline teletype controller had a one bit buffer per line so the computer had to scan characters in and out one bit at a time. It was only 110 baud, but still. R's, John PS: I still have the crumbling paper manuals for both. From jericho at attrition.org Mon Nov 15 19:45:27 2021 From: jericho at attrition.org (jericho) Date: Mon, 15 Nov 2021 21:45:27 -0600 (CST) Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> Message-ID: <5b29528-be4-f05c-4cc-9f57695b63dc@attrition.org> On Mon, 15 Nov 2021, touch--- via Internet-history wrote: : Thus we have no funds to pay for such a service, were one to exist. Note that archive.org does not run mailing lists AND its services require a per-year fee. They don?t say what happens when you stop paying, but in-perpituity escrows aren?t cheap to setup, even with initial funds. HOWEVER, note that such an service has other constraints, notably: : I?ve contacted many more typical archive-friendly places, including the Computer History Museum, but nobody currently runs a list intended to be archival, so none were able to assist. Such lists hosted at the IEEE ComSoc or ACM SIGCOMM societies have no archive capability either. The 'cheat' way to do that is to have the list archive somewhere, anywhere, via e.g. Mailman & Pipermail. Once that starts, you can trigger an archive.org 'save' request for one page. In time their spider will start to crawl and archive the content. It isn't "museum quality" in the manner speaking, but it will help ensure the content gets preserved in a public fashion. May also be worth talking to Jason Scott, who is an archivist that works at IA now. He was doing incredible work preserving computer history centered around security, hacking, and adjacent scenes. He has a soft spot for history like this and might be able to work some magic behind the scenes there. .b From johnl at iecc.com Mon Nov 15 19:45:36 2021 From: johnl at iecc.com (John Levine) Date: 15 Nov 2021 22:45:36 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <20211115225811.7jn3jtzzbldh2e2w@crankycanuck.ca> Message-ID: <20211116034537.9FB3A2FFFDB2@ary.qy> It appears that Andrew Sullivan via Internet-history said: >Hi, > >On Mon, Nov 15, 2021 at 02:34:22PM -0800, Dave Crocker via Internet-history wrote: >>archived under museum-quality conditions, with easy public access. > >Easy public access we have. Museum-quality conditions are not something, I think, that the Internet Society can guarantee at least at present. I dunno. I've visited the Computer Museum at its second home on the wharf in Boston and I've visited ISOC, and I'd say by those standards ISOC was at least museum quality. If people have contacts at the CHM at its new home in California, they might be persuaded to keep a copy of the archive, as they do for the RFC series. R's, John From brian.e.carpenter at gmail.com Mon Nov 15 19:55:34 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Tue, 16 Nov 2021 16:55:34 +1300 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <20211116034537.9FB3A2FFFDB2@ary.qy> References: <20211116034537.9FB3A2FFFDB2@ary.qy> Message-ID: <7d2a8f64-958b-c983-8217-e284d8c44518@gmail.com> Marc Weber is on this list and is, or was, associated with the CHM. Regards Brian On 16-Nov-21 16:45, John Levine via Internet-history wrote: > It appears that Andrew Sullivan via Internet-history said: >> Hi, >> >> On Mon, Nov 15, 2021 at 02:34:22PM -0800, Dave Crocker via Internet-history wrote: >>> archived under museum-quality conditions, with easy public access. >> >> Easy public access we have. Museum-quality conditions are not something, I think, that the Internet Society can guarantee at least at present. > > I dunno. I've visited the Computer Museum at its second home on the wharf in Boston and I've visited ISOC, > and I'd say by those standards ISOC was at least museum quality. > > If people have contacts at the CHM at its new home in California, they > might be persuaded to keep a copy of the archive, as they do for the > RFC series. > > R's, > John > From touch at strayalpha.com Mon Nov 15 19:55:48 2021 From: touch at strayalpha.com (touch at strayalpha.com) Date: Mon, 15 Nov 2021 19:55:48 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> Message-ID: To add a bit: - The ISOC is a fairly stable place, given its role in the Internet more stable than companies ?created? for archiving, AFAICT - It may be futile to expect anything more note that IRL, there is no such thihg as a permanent archive either museums come and go, as do cemeteries even endowments come and go The archives are available online; if anyone wants to contact an archival source and point them to those pages, they can be crawled and backed up without moving the list. Joe ? Joe Touch, temporal epistemologist www.strayalpha.com From touch at strayalpha.com Mon Nov 15 19:59:03 2021 From: touch at strayalpha.com (touch at strayalpha.com) Date: Mon, 15 Nov 2021 19:59:03 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <5b29528-be4-f05c-4cc-9f57695b63dc@attrition.org> References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> <5b29528-be4-f05c-4cc-9f57695b63dc@attrition.org> Message-ID: <9DCCC564-9CAC-4A20-A812-C89BB68DBC23@strayalpha.com> Hi, > On Nov 15, 2021, at 7:45 PM, jericho via Internet-history wrote: > > The 'cheat' way to do that is to have the list archive somewhere, > anywhere, via e.g. Mailman & Pipermail. There?s a URL at the bottom of every post that should suffice. > Once that starts, you can trigger > an archive.org 'save' request for one page. In time their spider will > start to crawl and archive the content. It isn't "museum quality" in the > manner speaking, but it will help ensure the content gets preserved in a > public fashion. It?s probably already happening then; I have the link on other web pages. But again, this isn?t ?museum quality? per se. Joe From jmamodio at gmail.com Mon Nov 15 20:02:17 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 22:02:17 -0600 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <9DCCC564-9CAC-4A20-A812-C89BB68DBC23@strayalpha.com> References: <17437e2e-2303-e966-4f4e-335bd575ace1@dcrocker.net> <5b29528-be4-f05c-4cc-9f57695b63dc@attrition.org> <9DCCC564-9CAC-4A20-A812-C89BB68DBC23@strayalpha.com> Message-ID: Is there anywhere a "virtual Internet Museum" where all materials about the history and evolution of the Internet are collected and archived ? Sort of key documents, pictures, videos, reports, manuals, etc ... it seems that it is all spread out on the net, often google helps ... Miss the days where everything was available on Finland's ftp server :-) Cheers Jorge On Mon, Nov 15, 2021 at 9:59 PM touch--- via Internet-history < internet-history at elists.isoc.org> wrote: > Hi, > > > On Nov 15, 2021, at 7:45 PM, jericho via Internet-history < > internet-history at elists.isoc.org> wrote: > > > > The 'cheat' way to do that is to have the list archive somewhere, > > anywhere, via e.g. Mailman & Pipermail. > > There?s a URL at the bottom of every post that should suffice. > > > Once that starts, you can trigger > > an archive.org 'save' request for one page. In time their spider will > > start to crawl and archive the content. It isn't "museum quality" in the > > manner speaking, but it will help ensure the content gets preserved in a > > public fashion. > > It?s probably already happening then; I have the link on other web pages. > > But again, this isn?t ?museum quality? per se. > > Joe > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From steve at shinkuro.com Mon Nov 15 20:14:48 2021 From: steve at shinkuro.com (Steve Crocker) Date: Mon, 15 Nov 2021 23:14:48 -0500 Subject: [ih] Six modems In-Reply-To: References: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> <0271C632-96CF-4605-B9FD-9F43040687E1@shinkuro.com> Message-ID: Jorge, Your comment about 6 modems got my attention. I had not paid a lot of attention to BBN report 1763. It was written at the very beginning of the Arpanet contract work in January 1969. By the time BBN released BBN 1822 in April or May 1969, the IMP was capable of handling up to four hosts. I believe this was a change from the original plan to have just one host on an IMP. Alex or others can comment more authoritatively, but it seems clear -- or at least plausible -- that seven places existed within the chassis to put interfaces. When the focus was on just one host, that left room for a maximum of six modems. As the design progressed, it seems they changed their design to a max of four hosts and a max of four modems, with a limit of seven for both. I also recall that as the interfaces became more varied -- host interface, distant host (DH) interface, very distant host (VDH) interface -- there were other constraints. IIRC, the number 19 figured into the equation, Each interface consumed some number of whatever the quantity was. I think the VDH interface consumed something like four units, but a regular host interface consumed only two. But I haven't checked the documentation and am not certain of the specifics. Steve On Mon, Nov 15, 2021 at 10:00 PM Jorge Amodio wrote: > > Weren't the BBN modems capable of handling up to 6 "phone lines" at > 50,000bps. So basically analog dedicated phone lines. > > I'm sure the early lines were not DDS 56K > > Regards > Jorge > > > On Mon, Nov 15, 2021 at 8:35 PM Steve Crocker via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Arpanet lines were 50 kbs, not 56 kbs. >> >> Sent from my iPhone >> >> > On Nov 15, 2021, at 9:11 PM, Timothy J. Salo via Internet-history < >> internet-history at elists.isoc.org> wrote: >> > >> > ?I think we all agree that the IMP was a pretty limited machine. From >> > the backup slides of a presentation of mine: >> > >> > Early ARPANET router, Interface Message Processor (IMP), (1969): >> > o 16-bit words, 12-16 K-word memory >> > o 100-?sec clock (10 KHz) >> > o Early ARPANET links: 56 kbps >> > o 0.18 clock cycles per bit >> > >> > I have argued that this, 12-16 K words of memory, is why we had the >> > end-to-end argument (which morphed into a principal and then into a >> > canon). >> > >> > (The rest of the presentation pretty much ignores the end-to-end >> > argument.) >> > >> > Also from this presentation: >> > >> > Early NSFNET router: DEC LSI-11/73 (1983) with Fuzzball router >> > o 512 KB memory >> > o (15.2 MHz) >> > o 271 clock cycles/bit >> > >> > -tjs >> > -- >> > Internet-history mailing list >> > Internet-history at elists.isoc.org >> > https://elists.isoc.org/mailman/listinfo/internet-history >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > From touch at strayalpha.com Mon Nov 15 20:16:21 2021 From: touch at strayalpha.com (Joe Touch) Date: Mon, 15 Nov 2021 20:16:21 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: Message-ID: > On Nov 15, 2021, at 8:02 PM, Jorge Amodio wrote: > > Is there anywhere a "virtual Internet Museum" where all materials about the history and evolution of the Internet are collected and archived ? Not really. Archive.org may aspire to be that, eventually, but it has no endowment yet, so doesn?t qualify as truly archival-capable IMO. And IRL even real, endowed museums don?t take all donations. It takes resources to maintain them - even digital ones. But rot, backup format conversion, and user format conversions all have real costs. And museums come and go. The computer history museum is in its third incarnation. Joe From tte at cs.fau.de Mon Nov 15 20:16:24 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Tue, 16 Nov 2021 05:16:24 +0100 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: <20211115215329.CAE4718C090@mercury.lcs.mit.edu> <713eefee-c573-d5a3-4699-e96d31178fbb@3kitty.org> <5631403d-9305-96ea-5ffa-ea883997a2b6@gmail.com> Message-ID: On Mon, Nov 15, 2021 at 07:19:15PM -0800, Tony Li via Internet-history wrote: > Even the DSP was insufficient, so it was assisted by additional FPGA based > processing and then (after a major wrong turn) things transitioned to an ASIC > based forwarding architecture. Wrong turn = 7200/7500 CPU forwarding ? > Microprocessors for all of their benefits have never been sufficient. More silicon has > always been a fundamental requirement. > > T > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From jmamodio at gmail.com Mon Nov 15 20:41:30 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Mon, 15 Nov 2021 22:41:30 -0600 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: Message-ID: I see, also if you want ?museum grade? it will require people with very good clue (paid+volunteers) to act as curators. -Jorge > On Nov 15, 2021, at 10:16 PM, Joe Touch wrote: > > ? >> On Nov 15, 2021, at 8:02 PM, Jorge Amodio wrote: >> >> Is there anywhere a "virtual Internet Museum" where all materials about the history and evolution of the Internet are collected and archived ? > > Not really. Archive.org may aspire to be that, eventually, but it has no endowment yet, so doesn?t qualify as truly archival-capable IMO. > > And IRL even real, endowed museums don?t take all donations. It takes resources to maintain them - even digital ones. But rot, backup format conversion, and user format conversions all have real costs. > > And museums come and go. The computer history museum is in its third incarnation. > > Joe From jack at 3kitty.org Mon Nov 15 21:43:20 2021 From: jack at 3kitty.org (Jack Haverty) Date: Mon, 15 Nov 2021 21:43:20 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <20211116025848.E3A4D18C090@mercury.lcs.mit.edu> References: <20211116025848.E3A4D18C090@mercury.lcs.mit.edu> Message-ID: <0b4c87a9-a4d5-fdff-fb55-8ef4cee101c5@3kitty.org> Yep, you're right that the IMP had I/O by DMA.?? But it wasn't just used for performance needs (those IMP guys were very clever...) It's well documented that the IMP had the ability to reload its program from a neighboring IMP, and even to deploy new releases of an IMP by successive reloads until the whole network was running the new release, and to do that without disrupting traffic flows.?? Kind of like open heart surgery...without losing the patient. As part of that patent dispute, I had to figure out *how* it did those things and describe it at an instruction-by-instruction level.? That's all in a report I submitted but I don't think it's publicly available (talk to the judge, not me).??? The ability to do those things relied on the behavior of the DMA, interrupt mechanisms, CPU, and some custom add-on hardware developed by BBN and/or Honeywell. The IMP custom hardware included timer functions, and in particular a "watchdog timer" which the program could start.?? If the timer ever ran down, the processor would get an unblockable interrupt and be forced to execute in the interrupt handler.?? That capability was used primarily to recover from software failures.?? If the program didn't reset the watchdog soon enough (it was in timeframes of seconds, not milliseconds), the interrupt would fire and force the PC to execuite the interrupt handling code.?? So the IMP program was designed to reset the watchdog frequently as part of its "main loop".??? If everything went well, the watchdog timer would never run out. If the watchdog interrupt handler ran, it would assume that something was seriously wrong, and initiate a reload of the entire IMP program from some other IMP.?? There were other ways to cause a reload from a neighbor.?? The network operator could command the IMP to reload.? Consistency checks performed periodically in the software, e.g., detecting that something that "should never happen" had in fact happened, would also be handled by jumping to the reload code.? There were quite a few situations which would trigger a reload; I probably only found some of them. Once the IMP was executing the reload code, it would block all interrupts.? There might still be I/O in progress, but when it completed nothing would happen.?? Instead, the IMP would pick one of its neighbors by random choice, prepare a "please reload me" packet, and issue an I/O write on the appropriate interface to have the DMA send that packet out the wire to the chosen neighbor.? It would then execute an I/O read instruction on that same interface, specifying that whatever came in on that interface next should be placed in its memory, overwriting the whole area in which the running IMP code was kept.?? Having done that, it would drop to an instruction that just did a single-instruction loop (i.e., a "JUMP "). The IMP would hang forever executing that instruction -- but the DMA and watchdog are still active.?? And the neighbor IMP is hopefully receiving the "reload me" request and then issuing an I/O write to send its entire code space of its memory out the wire to the other IMP which has an I/O read active to receive it and overwite its own whole code memory. Effectively the machine is now locked in a tight loop, with all interrupts disabled except for the one associated with that read operation ... and the watchdog timer.?? And the I/O is still happening. When the I/O interrupt finally fired, the handler would examine the pointers from the I/O to see if it finished cleanly, and if so JUMP to the starting address of the IMP code, pretty much as if it had just been powered on.?? If something looked amiss, e.g., the I/O transfer was too short to have delivered the whole IMP code space, the handler would pick another random neighbor IMP and go back to the reload code to get a memory image from that one instead.?? If the watchdog interrupt happened before the I/O completed, something was also wrong, and it would similarly try a different IMP to see if it could get a good reload.?? Today, it would probably also post a scathing Yelp review about that other IMP that failed to deliver. Of course, the code coming from a neighboring IMP might be a newer or older software release.?? Things in memory tend to move around when software gets revised, and the running IMP code was being overwritten by the DMA -- as it was running.?? So at some point, the instruction that the IMP was executing would be overwritten by whatever instruction lived at that memory address in the IMP code release coming from the neighbor.??? It seems that this was handled by programmer discipline, i.e., making sure that the "JUMP " instruction was present at the same memory address in every core image that might be in use in the ARPANET at the time. I surmised this when I found a small fragment of code that it appeared could never be executed.?? You couldn't get there from anywhere else in the code.?? But, in a *different release of the code* that's likely where the "JUMP " instruction was stored.?? So when the DMA overwrote the instruction the IMP was currently executing, the instruction didn't change at all, regardless of which releases were in the sending and receiving IMPs. When I figured this all out, I concluded that the IMP designers were indeed very clever.?? But as far as I know, what I wrote above has never been documented elsewhere.? Not even as comments in the code. So, when I suggested that it was important to look at more than CPU speed, memory capacity, et al, this is the kind of functionality I was thinking about.? Steve's question about a microprocessor being able to do what an IMP did means being able to somehow implement similar I/O, timer, interrupt, DMA, and any other "clever" behavior of the H516-based IMP, using some combination of add-on hardware (I/O, timer etc.) and an instruction set and processor behavior that supported whatever "clever programming" was needed to get functionality such as the reload and/or software upgrade capability, and perhaps other such stuff I didn't encounter since I only looked at one aspect of the IMP code. Could the 4004 do it??? I have no idea, someone who knows a lot more about ancient microprocessors would have to figure out how they might do what the IMP did. Enjoy, /Jack Haverty On 11/15/21 6:58 PM, Noel Chiappa via Internet-history wrote: > > E.g. a register-register MOV took 1.7 microseconds, but a memory-memory > > move .. seems to have taken 1.7+4.9=6.6 usec, so about 1/5th of a MIP > > Ooops, I mistakenly was in the -11/23 timing appendix, not the -11/03; the > latter was considerably slower: 3.5 usec basic, 2.5(sic)+9.1=11.6 usec. > (Indirect was even slower.) So about 1/10th of a MIP. > > > > From: Jack Haverty > > > E.g., to handle I/O on a serial interface, the CPU might have to take > > an interrupt on every byte > > ... > > I think that kind of issue is the one Alex referred to about selecting > > the 316 because of its interrupt mechanism. ... how the hardware > > handled I/O on all those interfaces was a crucial factor in selecting > > the 516. > > Yes and no. The IMP's modem and host interfaces were both DMA (in the sense > that the CPU only got a single interrupt - which diverted instruction > processing in the CPU to other instructions - for every packet, not on every > word; the details differed significantly from modern DMA, though - see > below). But there was a timing issue. > > Per 'The interface message processor for the ARPA computer network', > available here: > > https://www.walden-family.com/public/1970-imp-afips.pdf > > "To send a packet, the IMP program sets up memory pointers to the packet and > then activates the interface ... The interface takes successive words from > the memory using its assigned output data channel and transmits them > bit-serially (to the Host or to the modem). When the memory buffer has thus > been emptied, the interface notifies the program via an interrupt". > > The details are intetesting: the IMP used "a set of 16 multiplexed channels > (which implement a 4-cycle data break)". This is through a device called the > DMC, the 'Direct Multiplex Control' (sometimes 'Data Multiplex Control'). > Notice the "4-cycle data break"; consulting the DMC manual (also online), > this was very similar to the '3-cycle data break' used on many early DEC > machines, up through the PDP-8. This kept the buffer address and count in > main memory (to reduce to cost of devices); the downside is that it increased > the memory bandwidth usage. (The 4 cycles were 1) current buffer address > read, ii) buffer extent read, iii) data read/write, iv) modified buffer > address write-back.) > > (Hey, it could have been worse; the DM11 asynchronous line interface of the > early PDP-11 kept the _shift registers_ in main memory, and used DMA to gain > access to them during input/output. At least it was efficient DMA - the > memory address was stored in the device! :-) > > The timing issue comes from the fact that, as far as I can tell from > the IMP hardware manual: > > https://walden-family.com/impcode/imp-hardware.pdf > > there was _no_ buffering on the modem and host interfaces, just the shift > register; not so bad on the host interface, which used a handshake, and could > be paused, but potentially problematic on the synchronous modem interface; > after a word arrived, it had to be written to memory before the first _bit_ > of the next word arrived. (The DM11 had the same issue.) > > > > That's why I suggested that the I/O capabilities of a microprocessor > > needed to be considered when trying to figure out how it compared to > > the 516, more so than just classic metrics like raw memory and CPU speed. > > That part I agree with. (But don't forget the address space, either; the > 4004 really had too small an address space to be usable as a router at _any > point in time_.) > > Noel From lars at nocrew.org Mon Nov 15 23:13:43 2021 From: lars at nocrew.org (Lars Brinkhoff) Date: Tue, 16 Nov 2021 07:13:43 +0000 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <934040ee-2d99-2689-2a41-d8ccf35a18a7@3kitty.org> (Jack Haverty via Internet-history's message of "Mon, 15 Nov 2021 19:09:06 -0800") References: <934040ee-2d99-2689-2a41-d8ccf35a18a7@3kitty.org> Message-ID: <7wzgq4vahk.fsf@junk.nocrew.org> Jack Haverty wrote: > There is quite a lot of information about the IMP code, including > several complete listings of its program, in 316 assembler of course, > and its recent resurrection, available at: > https://walden-family.com/impcode/ > > I'm not sure where it is now, but the code itself also exists in > runnable form A ready-to-go sample configuration is distributed with the SIMH emulator: https://github.com/simh/simh/tree/master/H316/tests > It's ready for someone to resurrect one of those early host computers > to create traffic. I'm trying to do that, but finding suitable host systems is not that easy. NCP implementations are hard to find now. On top of that the IMP code is from 1973, so it will only accept the old 32-bit leader format. From jeanjour at comcast.net Tue Nov 16 05:03:10 2021 From: jeanjour at comcast.net (John Day) Date: Tue, 16 Nov 2021 08:03:10 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> Message-ID: <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> We have no means of archival store for digital material. Archival has to last 500 years or more. (I am often using materials that old.) Not only are we not sure the storage media last that long, but having something that reads it is the real problem.) > On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history wrote: > >>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history wrote: >>> I got a private note asking about a paid archive at archive.org. I think that a reasonable question and possibly worth doing on it own. Whether that qualifies as 'museum quality' isn't something I can assess. > > Well, it?s what we?ve got for ?archival quality? on the Internet right now, and if it?s not good enough, that means Brewster needs more support. > > -Bill > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From vgcerf at gmail.com Tue Nov 16 06:28:57 2021 From: vgcerf at gmail.com (vinton cerf) Date: Tue, 16 Nov 2021 09:28:57 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> Message-ID: You are singing my song V On Tue, Nov 16, 2021, 08:03 John Day via Internet-history < internet-history at elists.isoc.org> wrote: > We have no means of archival store for digital material. Archival has to > last 500 years or more. (I am often using materials that old.) Not only are > we not sure the storage media last that long, but having something that > reads it is the real problem.) > > > On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history < > internet-history at elists.isoc.org> wrote: > > > >>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history < > internet-history at elists.isoc.org> wrote: > >>> I got a private note asking about a paid archive at archive.org. I > think that a reasonable question and possibly worth doing on it own. > Whether that qualifies as 'museum quality' isn't something I can assess. > > > > Well, it?s what we?ve got for ?archival quality? on the Internet right > now, and if it?s not good enough, that means Brewster needs more support. > > > > -Bill > > > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From jmamodio at gmail.com Tue Nov 16 07:10:43 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Tue, 16 Nov 2021 09:10:43 -0600 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> Message-ID: Agreed, but who is "we" ? I think that if funds and bodies are available, on of the first steps after having some sort of institutional framework should be collecting/acquiring and curating materials and provide temporary storage while some working group starts to work on a permanent long term solution. I agree it is not an easy task and there are probably many variables but just talking it can't be done only means it will never exist. Cheers Jorge On Tue, Nov 16, 2021 at 7:03 AM John Day via Internet-history < internet-history at elists.isoc.org> wrote: > We have no means of archival store for digital material. Archival has to > last 500 years or more. (I am often using materials that old.) Not only are > we not sure the storage media last that long, but having something that > reads it is the real problem.) > > > On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history < > internet-history at elists.isoc.org> wrote: > > > >>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history < > internet-history at elists.isoc.org> wrote: > >>> I got a private note asking about a paid archive at archive.org. I > think that a reasonable question and possibly worth doing on it own. > Whether that qualifies as 'museum quality' isn't something I can assess. > > > > Well, it?s what we?ve got for ?archival quality? on the Internet right > now, and if it?s not good enough, that means Brewster needs more support. > > > > -Bill > > > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From dhc at dcrocker.net Tue Nov 16 07:34:19 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Tue, 16 Nov 2021 07:34:19 -0800 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> Message-ID: <17890cea-ac85-3642-e728-7cdc480b1363@dcrocker.net> On 11/16/2021 7:10 AM, Jorge Amodio wrote: > I think that if funds and bodies are available, on of the first steps > after having some sort of institutional framework should be > collecting/acquiring and curating materials and provide temporary > storage while some working group starts to work on a permanent long term > solution. Thanks for writing that. It points in the direction I'm hoping we can go. Perhaps it will help to have a fairly simple summary statement of the steady-state goals for this? I said "museum quality" as a facile way to distinguish from simple backup or normal computer archive. But what are the technical and operational requirements? The reference to 500 years is helpful. This probably needs to levels of statement. The first is a summary of affirmative goals and the nature of threats/problems to protect against. The second might be a statement of technical approaches considered most promising to aid in sustaining the goals. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From ajs at crankycanuck.ca Tue Nov 16 07:54:54 2021 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Tue, 16 Nov 2021 10:54:54 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> Message-ID: <20211116155454.ojv6daq6wdwvjjbr@crankycanuck.ca> On Mon, Nov 15, 2021 at 09:33:22PM -0500, Andrew Sullivan via Internet-history wrote: > >We do archive the Internet Society pages at archive.org. I am not sure whether that extends to list archives but I will find out. I looked into this, and we are not currently including mailing list archives in the service we pay archive-it for. I'm also not totally sure adding it would really help, because of the "museum quality" notion -- we do have an archive, already, and putting up two archives on best-efforts basis doesn't really meet that criterion. The Internet Society, of course, is committed to the continued support of the list and its archives, and I don't anticipate any diminution in that commitment. But clearly, if there were a mechanism for long-term archival storage of valuable online resources, I think we would be interested in feeding it. Hope that helps, A (with my ISOC hat on) -- Andrew Sullivan ajs at crankycanuck.ca From wfms at wfms.org Tue Nov 16 08:18:38 2021 From: wfms at wfms.org (wfms at wfms.org) Date: Tue, 16 Nov 2021 16:18:38 +0000 (UTC) Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> Message-ID: <2844a159-8a5a-0fa-d9fd-b98ae3189632@wfms.org> I had quite an interesting discussion while sorting through some materials (mostly digital) with a profesional (non-techie) archivist. After quite discussion over what MTBF, digital storage methods, longevities of various media, etc. it got to the point where we both looked around...lots of boxes full of pulp - some of it going back 2-3 centuries, still accessible. We found a printer. To get fancier, could have made a preservation copy as well as a consultation copy. Even fancier, sent the 8.5" x 14" off for binding. That's not to say paper-bound i or isn't a step backwards, and nothing beats feeding digital data into a search engine indexer - but at least some semblance of indexing into the various volumes also not paper-bound may help. It all depends. On Tue, 16 Nov 2021, vinton cerf via Internet-history wrote: > You are singing my song > V > > On Tue, Nov 16, 2021, 08:03 John Day via Internet-history < > internet-history at elists.isoc.org> wrote: > >> We have no means of archival store for digital material. Archival has to >> last 500 years or more. (I am often using materials that old.) Not only are >> we not sure the storage media last that long, but having something that >> reads it is the real problem.) >> >>> On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history < >> internet-history at elists.isoc.org> wrote: >>> >>>>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history < >> internet-history at elists.isoc.org> wrote: >>>>> I got a private note asking about a paid archive at archive.org. I >> think that a reasonable question and possibly worth doing on it own. >> Whether that qualifies as 'museum quality' isn't something I can assess. >>> >>> Well, it?s what we?ve got for ?archival quality? on the Internet right >> now, and if it?s not good enough, that means Brewster needs more support. >>> >>> -Bill >>> >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history >> >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > wfms From jeanjour at comcast.net Tue Nov 16 08:34:57 2021 From: jeanjour at comcast.net (John Day) Date: Tue, 16 Nov 2021 11:34:57 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <2844a159-8a5a-0fa-d9fd-b98ae3189632@wfms.org> References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> <2844a159-8a5a-0fa-d9fd-b98ae3189632@wfms.org> Message-ID: <2901F2BC-3CB6-4261-A716-D4E78785A2F2@comcast.net> I have come to the same conclusion for the same reasons, with the same regrets. ;-) > On Nov 16, 2021, at 11:18, wfms at wfms.org wrote: > > > I had quite an interesting discussion while sorting through some materials (mostly digital) with a profesional (non-techie) archivist. > > After quite discussion over what MTBF, digital storage methods, longevities of various media, etc. it got to the point where we both looked around...lots of boxes full of pulp - some of it going back 2-3 centuries, still accessible. > > We found a printer. > > To get fancier, could have made a preservation copy as well as a consultation copy. Even fancier, sent the 8.5" x 14" off for binding. > > That's not to say paper-bound i or isn't a step backwards, and nothing beats feeding digital data into a search engine indexer - but at least some semblance of indexing into the various volumes also not paper-bound may help. It all depends. > > On Tue, 16 Nov 2021, vinton cerf via Internet-history wrote: > >> You are singing my song >> V >> >> On Tue, Nov 16, 2021, 08:03 John Day via Internet-history < >> internet-history at elists.isoc.org> wrote: >> >>> We have no means of archival store for digital material. Archival has to >>> last 500 years or more. (I am often using materials that old.) Not only are >>> we not sure the storage media last that long, but having something that >>> reads it is the real problem.) >>> >>>> On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history < >>> internet-history at elists.isoc.org> wrote: >>>> >>>>>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history < >>> internet-history at elists.isoc.org> wrote: >>>>>> I got a private note asking about a paid archive at archive.org. I >>> think that a reasonable question and possibly worth doing on it own. >>> Whether that qualifies as 'museum quality' isn't something I can assess. >>>> >>>> Well, it?s what we?ve got for ?archival quality? on the Internet right >>> now, and if it?s not good enough, that means Brewster needs more support. >>>> >>>> -Bill >>>> >>>> -- >>>> Internet-history mailing list >>>> Internet-history at elists.isoc.org >>>> https://elists.isoc.org/mailman/listinfo/internet-history >>> >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history >>> >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > > > wfms From jnc at mercury.lcs.mit.edu Tue Nov 16 09:16:57 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Tue, 16 Nov 2021 12:16:57 -0500 (EST) Subject: [ih] Intel 4004 vs the IMP Message-ID: <20211116171657.C959318C090@mercury.lcs.mit.edu> > From: Jack Haverty > in particular a "watchdog timer" The TIU (BTW, I have an announcement about the recovery of its source and documentation, which I'll get to after the list quiets down) also had a watchdog timer board. (We used those boards on the MIT routers, with different code; I'm not sure if anyone else did, e.g. BBN QBUS routers. The idea was carried over to the Proteon p4200, and probably still exists in other routers today.) The TIU one had a less restricted operating environment, since the PDP-11 reserved a small block of address space for bootstrap ROM, and the machine could be set to jump to that on reset. (For larger bootstrap codes, PDP-11's often hid the larger code in ROM behind a 'window', and the initial code copies it out to main memory to run it. I can't remember if the TIU one did that.) Which shows again the need to carefully examine all areas of a target machine's functionality to assess its suitability. > So at some point, the instruction that the IMP was executing would be > overwritten by whatever instruction lived at that memory address in the > IMP code release coming from the neighbor. It seems that this was > handled by programmer discipline The IMP custom hardware documentation: https://walden-family.com/impcode/imp-hardware.pdf indicates (page 4-6) that it had the ability to write-protect protect the watchdog interrupt vector, and also 'sector 1' of main memory (which presumably, said vector could be set to point to). I'm not sure if this capability was also utilized to make reload more robust (i.e. to protect the 'JMP .') Certainly the IMPs were enormously robust against 'brain freeze'; in several years as the liaison for the MIT IMPs, I don't recall ever having to manually restart one. > Could the 4004 do it I think I've already answered this: 'probably not'. (And as an implementer of early routers, I think I have the detailed knowledge to make this assessment.) To re-state my reasoning, which apparently got missed, the maximum amount of RAM it could address was about 650 bytes - enough for only a single packet buffer. I don't think one could build a viable router with only a single packet buffer: e.g. if a packet started arriving on one interface, all the others would have to be shut off. Although one could probably have bodged up some sort of bank switching hardware arrangement which allowed the CPU to only 'see' a single packet buffer at a time, to partially get around that. (The PDP-11/23 variant of the CGW did exactly that, so that approach worked - including as the main Stanford ARPANET gateway for some years - although it kind of depended on having DMA network interfaces which could 'see' all the packet buffers all the time; supporting a DMA interface which could only 'see' the bottom 64KB of memory took some contortions in the code.) Non-DMA interfaces, where the CPU had to get invlved in moving data, would similarly have been an issue - lots of bank swithing going on; but such interfaces would have been throughput bottlenecks _anyway_. It would have depended on what the performance goals were, what kind of monetary budget one had for add-on work-around hardware ('an engineer is someone who can do for $.5 what any fool can do for $2'), etc. Also, it's not worth looking too hard at the 4004, when the 8008 (available _very_ shortly after the 4004) was considerably more capable. (I never worked with 8008's BITD, but the programmer's front panel for the PDP-11/04&34 used them, and in debugging some panels recently I read a lot of the code, so I now know a tiny bit about them.) Their address space was still pretty small (16KB, but I'm not sure if that was ROM and RAM, or just ROM, like the 4004), but it was probably enough for a rudimentary router. Especially if one could spend on bank-switching, which was used very sucessfully in a number of early DEC machines, including the very successful PDP-8 (from the original 'straight 8', all the way on). Noel From amckenzie3 at yahoo.com Tue Nov 16 09:39:35 2021 From: amckenzie3 at yahoo.com (Alex McKenzie) Date: Tue, 16 Nov 2021 17:39:35 +0000 (UTC) Subject: [ih] Intel 4004 vs the IMP In-Reply-To: References: <91a3618c-6f23-37d3-6715-761f56b28109@saloits.com> <0271C632-96CF-4605-B9FD-9F43040687E1@shinkuro.com> Message-ID: <246582929.862558.1637084375996@mail.yahoo.com> The IMP could have up to 7 connected devices total.? Up to 5 could be phone lines.? Up to 4 could be hosts. The terminal handling portion of the TIP counted as a host.? [When the Very Distant Host Interface was introduced in the mid-1970's it counted as 2 devices.]? There was one ARPAnet circuit that ran at 230kbps (within NASA AMES) and two that ran at 19.2kbps (US to Norway, and Norway to England).? As far as I can recall, all the rest ran at 50kbps. Cheers,Alex On Monday, November 15, 2021, 10:01:05 PM EST, Jorge Amodio via Internet-history wrote: Weren't the BBN modems capable of handling up to 6 "phone lines" at 50,000bps. So basically analog dedicated phone lines. I'm sure the early lines were not DDS 56K Regards Jorge On Mon, Nov 15, 2021 at 8:35 PM Steve Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > Arpanet lines were 50 kbs, not 56 kbs. > > Sent from my iPhone > > > On Nov 15, 2021, at 9:11 PM, Timothy J. Salo via Internet-history < > internet-history at elists.isoc.org> wrote: > > > > ?I think we all agree that the IMP was a pretty limited machine.? From > > the backup slides of a presentation of mine: > > > > Early ARPANET router, Interface Message Processor (IMP), (1969): > > o 16-bit words, 12-16 K-word memory > > o 100-?sec clock (10 KHz) > > o Early ARPANET links: 56 kbps > > o 0.18 clock cycles per bit > > > > I have argued that this, 12-16 K words of memory, is why we had the > > end-to-end argument (which morphed into a principal and then into a > > canon). > > > > (The rest of the presentation pretty much ignores the end-to-end > > argument.) > > > > Also from this presentation: > > > > Early NSFNET router: DEC LSI-11/73 (1983) with Fuzzball router > > o 512 KB memory > > o (15.2 MHz) > > o 271 clock cycles/bit > > > > -tjs > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From jmamodio at gmail.com Tue Nov 16 11:44:26 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Tue, 16 Nov 2021 13:44:26 -0600 Subject: [ih] Museum-quality archive for this list? In-Reply-To: <2844a159-8a5a-0fa-d9fd-b98ae3189632@wfms.org> References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> <2844a159-8a5a-0fa-d9fd-b98ae3189632@wfms.org> Message-ID: Preserving paper for 500+ years will require an environmentally controlled room, and for the amount of documents associated, a very large room ! -J On Tue, Nov 16, 2021 at 10:19 AM wfms--- via Internet-history < internet-history at elists.isoc.org> wrote: > > I had quite an interesting discussion while sorting through some materials > (mostly digital) with a profesional (non-techie) archivist. > > After quite discussion over what MTBF, digital storage methods, > longevities of various media, etc. it got to the point where we both > looked around...lots of boxes full of pulp - some of it going back 2-3 > centuries, still accessible. > > We found a printer. > > To get fancier, could have made a preservation copy as well as a > consultation copy. Even fancier, sent the 8.5" x 14" off for binding. > > That's not to say paper-bound i or isn't a step backwards, and nothing > beats feeding digital data into a search engine indexer - but at least > some semblance of indexing into the various volumes also not paper-bound > may help. It all depends. > > On Tue, 16 Nov 2021, vinton cerf via Internet-history wrote: > > > You are singing my song > > V > > > > On Tue, Nov 16, 2021, 08:03 John Day via Internet-history < > > internet-history at elists.isoc.org> wrote: > > > >> We have no means of archival store for digital material. Archival has to > >> last 500 years or more. (I am often using materials that old.) Not only > are > >> we not sure the storage media last that long, but having something that > >> reads it is the real problem.) > >> > >>> On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history < > >> internet-history at elists.isoc.org> wrote: > >>> > >>>>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history < > >> internet-history at elists.isoc.org> wrote: > >>>>> I got a private note asking about a paid archive at archive.org. I > >> think that a reasonable question and possibly worth doing on it own. > >> Whether that qualifies as 'museum quality' isn't something I can assess. > >>> > >>> Well, it?s what we?ve got for ?archival quality? on the Internet right > >> now, and if it?s not good enough, that means Brewster needs more > support. > >>> > >>> -Bill > >>> > >>> -- > >>> Internet-history mailing list > >>> Internet-history at elists.isoc.org > >>> https://elists.isoc.org/mailman/listinfo/internet-history > >> > >> -- > >> Internet-history mailing list > >> Internet-history at elists.isoc.org > >> https://elists.isoc.org/mailman/listinfo/internet-history > >> > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > > > > wfms > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From jack at 3kitty.org Tue Nov 16 12:26:58 2021 From: jack at 3kitty.org (Jack Haverty) Date: Tue, 16 Nov 2021 12:26:58 -0800 Subject: [ih] Intel 4004 vs the IMP In-Reply-To: <66237FBB-00B9-4414-9F88-149EE7C7F2BE@bromirski.net> References: <20211116025848.E3A4D18C090@mercury.lcs.mit.edu> <0b4c87a9-a4d5-fdff-fb55-8ef4cee101c5@3kitty.org> <66237FBB-00B9-4414-9F88-149EE7C7F2BE@bromirski.net> Message-ID: Not right, the IMP did all I/O using DMA, including normal packet traffic as well as the code image download.?? The code image was of course much larger than the usual packet I/O.?? I don't recall all the exact behavior now, but basically the IMP needing code reload would abort all I/O, issue a write instruction to send out the reload request to a neighbor, and execute a read instruction to put whatever subsequently arrived on that interface from the neighbor to replace the contents of its program space in memory.?? Then it waited in a single instruction loop for the interrupt on completion of the read or expiration of the watchdog timer. The IMP being reloaded would stop handling traffic.?? In the typical case where something was "seriously wrong", it probably had already stopped handling regular traffic.?? All other IMPs kept running normally, so the ARPANET kept moving user traffic for everyone else. Essentially the IMP was a multiprocessor, with the main CPU as well as I/O using DMA.? It was a bit challenging to mentally trace the program path over time since some of the "state" of the machine was maintained by modifying instructions in the active program.?? So when looking for example at how the IMP handled some interrupt, I couldn't assume that the instructions shown in the listing were actually what was in memory at the time when the interrupt handler ran.?? The instructions might have been changed earlier, so when the handler ran it did something different from what you'd expect by simply looking at the code listing.?? Some instructions in the program were essentially "state variables" of the overall IMP program design. Jack Haverty On 11/16/21 4:58 AM, Lukasz Bromirski wrote: > Jack, > >> On 16 Nov 2021, at 06:43, Jack Haverty via Internet-history wrote: >> >> Once the IMP was executing the reload code, it would block all interrupts. There might still be I/O in progress, but when it completed nothing would happen. [?] >> The IMP would hang forever executing that instruction -- but the DMA and watchdog are still active. And the neighbor IMP is hopefully receiving the "reload me" request and then issuing an I/O write to send its entire code space of its memory out the wire to the other IMP which has an I/O read active to receive it and overwite its own whole code memory. >> Effectively the machine is now locked in a tight loop, with all interrupts disabled except for the one associated with that read operation ... and the watchdog timer. And the I/O is still happening. > So given the description, IMP would stop forwarding traffic during such > remote reload ? or not? The I/O and DMA statements relate to ONLY > serving code download, right? > From jeanjour at comcast.net Tue Nov 16 13:37:59 2021 From: jeanjour at comcast.net (John Day) Date: Tue, 16 Nov 2021 16:37:59 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> <2844a159-8a5a-0fa-d9fd-b98ae3189632@wfms.org> Message-ID: <163F1447-A18B-404E-851C-71081F6E4F56@comcast.net> Not really. You wouldn?t believe the conditions I have found paper! The stacks of the Vatican Library are not temperature controlled, in fact no controls at all. That is also true of the old Bibliotheque Nationale in Paris and the British Library. Yet the stuff is there and in use. Heck, I was shown 2 400 year old maps hanging in a junk room under the Vatican Library no protection at all. We have paper dating back millennia. > On Nov 16, 2021, at 14:44, Jorge Amodio via Internet-history wrote: > > Preserving paper for 500+ years will require an environmentally controlled > room, and for the amount of documents associated, a very large room ! > > -J > > On Tue, Nov 16, 2021 at 10:19 AM wfms--- via Internet-history < > internet-history at elists.isoc.org> wrote: > >> >> I had quite an interesting discussion while sorting through some materials >> (mostly digital) with a profesional (non-techie) archivist. >> >> After quite discussion over what MTBF, digital storage methods, >> longevities of various media, etc. it got to the point where we both >> looked around...lots of boxes full of pulp - some of it going back 2-3 >> centuries, still accessible. >> >> We found a printer. >> >> To get fancier, could have made a preservation copy as well as a >> consultation copy. Even fancier, sent the 8.5" x 14" off for binding. >> >> That's not to say paper-bound i or isn't a step backwards, and nothing >> beats feeding digital data into a search engine indexer - but at least >> some semblance of indexing into the various volumes also not paper-bound >> may help. It all depends. >> >> On Tue, 16 Nov 2021, vinton cerf via Internet-history wrote: >> >>> You are singing my song >>> V >>> >>> On Tue, Nov 16, 2021, 08:03 John Day via Internet-history < >>> internet-history at elists.isoc.org> wrote: >>> >>>> We have no means of archival store for digital material. Archival has to >>>> last 500 years or more. (I am often using materials that old.) Not only >> are >>>> we not sure the storage media last that long, but having something that >>>> reads it is the real problem.) >>>> >>>>> On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history < >>>> internet-history at elists.isoc.org> wrote: >>>>> >>>>>>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history < >>>> internet-history at elists.isoc.org> wrote: >>>>>>> I got a private note asking about a paid archive at archive.org. I >>>> think that a reasonable question and possibly worth doing on it own. >>>> Whether that qualifies as 'museum quality' isn't something I can assess. >>>>> >>>>> Well, it?s what we?ve got for ?archival quality? on the Internet right >>>> now, and if it?s not good enough, that means Brewster needs more >> support. >>>>> >>>>> -Bill >>>>> >>>>> -- >>>>> Internet-history mailing list >>>>> Internet-history at elists.isoc.org >>>>> https://elists.isoc.org/mailman/listinfo/internet-history >>>> >>>> -- >>>> Internet-history mailing list >>>> Internet-history at elists.isoc.org >>>> https://elists.isoc.org/mailman/listinfo/internet-history >>>> >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history >>> >> >> >> wfms >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From jeanjour at comcast.net Tue Nov 16 13:41:25 2021 From: jeanjour at comcast.net (John Day) Date: Tue, 16 Nov 2021 16:41:25 -0500 Subject: [ih] Museum-quality archive for this list? In-Reply-To: References: <9d29e35f-42b4-4346-30fb-cbd3b4ce1893@dcrocker.net> <0F638A4B-2FA1-475A-9C19-4275EB533C50@crankycanuck.ca> <3AE26880-FF3B-4696-85B2-4FBD931A7C05@comcast.net> <2844a159-8a5a-0fa-d9fd-b98ae3189632@wfms.org> Message-ID: <2C02B75F-6B88-4B45-BD34-0B01F4D40727@comcast.net> Of course, it is nice if it can be. That is why I like the caverns under the library where the Charles Babbage Institute is. https://www.minnpost.com/stroll/2015/10/subterranean-caverns-protect-us-andersen-library-collections/ They are ~ 40 x 60 x 600 feet. > On Nov 16, 2021, at 14:44, Jorge Amodio via Internet-history wrote: > > Preserving paper for 500+ years will require an environmentally controlled > room, and for the amount of documents associated, a very large room ! > > -J > > On Tue, Nov 16, 2021 at 10:19 AM wfms--- via Internet-history < > internet-history at elists.isoc.org> wrote: > >> >> I had quite an interesting discussion while sorting through some materials >> (mostly digital) with a profesional (non-techie) archivist. >> >> After quite discussion over what MTBF, digital storage methods, >> longevities of various media, etc. it got to the point where we both >> looked around...lots of boxes full of pulp - some of it going back 2-3 >> centuries, still accessible. >> >> We found a printer. >> >> To get fancier, could have made a preservation copy as well as a >> consultation copy. Even fancier, sent the 8.5" x 14" off for binding. >> >> That's not to say paper-bound i or isn't a step backwards, and nothing >> beats feeding digital data into a search engine indexer - but at least >> some semblance of indexing into the various volumes also not paper-bound >> may help. It all depends. >> >> On Tue, 16 Nov 2021, vinton cerf via Internet-history wrote: >> >>> You are singing my song >>> V >>> >>> On Tue, Nov 16, 2021, 08:03 John Day via Internet-history < >>> internet-history at elists.isoc.org> wrote: >>> >>>> We have no means of archival store for digital material. Archival has to >>>> last 500 years or more. (I am often using materials that old.) Not only >> are >>>> we not sure the storage media last that long, but having something that >>>> reads it is the real problem.) >>>> >>>>> On Nov 15, 2021, at 21:53, Bill Woodcock via Internet-history < >>>> internet-history at elists.isoc.org> wrote: >>>>> >>>>>>> On Nov 15, 2021, at 20:37, Dave Crocker via Internet-history < >>>> internet-history at elists.isoc.org> wrote: >>>>>>> I got a private note asking about a paid archive at archive.org. I >>>> think that a reasonable question and possibly worth doing on it own. >>>> Whether that qualifies as 'museum quality' isn't something I can assess. >>>>> >>>>> Well, it?s what we?ve got for ?archival quality? on the Internet right >>>> now, and if it?s not good enough, that means Brewster needs more >> support. >>>>> >>>>> -Bill >>>>> >>>>> -- >>>>> Internet-history mailing list >>>>> Internet-history at elists.isoc.org >>>>> https://elists.isoc.org/mailman/listinfo/internet-history >>>> >>>> -- >>>> Internet-history mailing list >>>> Internet-history at elists.isoc.org >>>> https://elists.isoc.org/mailman/listinfo/internet-history >>>> >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history >>> >> >> >> wfms >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From jack at 3kitty.org Mon Nov 22 12:50:28 2021 From: jack at 3kitty.org (Jack Haverty) Date: Mon, 22 Nov 2021 12:50:28 -0800 Subject: [ih] "The First Router" on Jeopardy Message-ID: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Watching Jeopardy from November 17, 2021, I was surprised to see a "clue" which was a picture of Len Kleinrock standing in front of the ARPANET IMP which has been preserved at UCLA.? It was a clue under the category "It's a New Machine". The host read the clue: "In 1969, Leonard Kleinrock and his team sent the first message over the Internet with the first THIS, which now connects devices like modems." None of the contestants knew the answer, although one guessed "backbone", which isn't a bad guess.? So the guest revealed the answer: "You're looking at the first router." My immediate reaction was "No, you're not!??? That's an IMP." See https://www.youtube.com/watch?v=Lmv1WKMGcDw starting at about 1:29 That's not quite like I remember it.? Ginny Strazisar built the first switching device for the Internet to connect the ARPANET to the Packet Radio net, circa 1977.? To me that was the genesis of the "Internet" - INTERconnecting of NETworks of a variety of types, using TCP/IP to glue it all together.?? But millions of people just learned otherwise. A historical tidbit -- Back in 1977 Ginny's system was called a "gateway", but later was renamed a "router".?? It's possible that I did the renaming.? At BBN we were selling lots of packet switches, and sometimes customers asked for ideas on how to use their LANs in the network.? Our sales people would tell them about the research activities, and the role of gateways, TCP/IP , and the Internet. But in many customers' minds that term "gateway"? immediately set off alarm bells, because they had prior bad experience with "gateways" in their IBM networks, and didn't want anything to do with more "gateways".?? So I suggested calling them "routers" instead of "gateways", and suddenly the marketplace became much more willing to listen. Enjoy, Jack Haverty From dhc at dcrocker.net Mon Nov 22 13:14:03 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Mon, 22 Nov 2021 13:14:03 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: <6769fd88-804c-b41f-9878-ff5eeb340e40@dcrocker.net> On 11/22/2021 12:50 PM, Jack Haverty via Internet-history wrote: > > My immediate reaction was "No, you're not!??? That's an IMP." Welcome to the distinction between a popular perception versus a professional one. The former lacks nuance, using very coarse metrics. The latter ought to be more refined, and sometimes is. For the world at large, packet switching is really the defining moment. For them, the moment is the invention of computer networking, rather than the invention of linking networks together. (That is, assuming that are careful enough to avoid confusing Web with Internet...[*]) Sometimes, the error is in the later direction. Getting even professionals to be careful in talking about email history is difficult. So it is quite common even in highly technical circles -- such as a couple of weeks ago for a press release -- for folk to say that Ray invented email rather than Ray invented networked email. d/ [*] maybe 20 years ago, taking a Spanish course in Spain, with a class of much (much) younger folk from all over Europe, the instructor prompted some discussion in Spanish by asking us about our backgrounds. I chose to say that I worked on the UCLA networking project, in 1972, explaining it was the first site on the Internet. One of the very bright, very young students objected vigorously, saying that the Internet was invented in 1989.[**] I smiled and tried to explain the difference but she persisted. The instructor didn't care about the answer, as long as everyone was talking in Spanish, but this dragged on. The youngster would not relent. Finally the instructor intervened, say "Look, you weren't born yet and he was there!" [**] I have heard of a name for it, but there should be one that distinguishes errors that require a lot of knowledge to make. That she knew of 1989 in the net's history was impressive. My first encounter with this type of error was while at the University of Delaware, around 1980, talking to a hotel reservation agent in Toronto. She asked for my address and when I said Newark, Delaware, she queried "that's a suburb of Philadelphia, isn't it?" sigh. -- Dave Crocker Brandenburg InternetWorking bbiw.net From vgcerf at gmail.com Mon Nov 22 13:34:36 2021 From: vgcerf at gmail.com (vinton cerf) Date: Mon, 22 Nov 2021 16:34:36 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: your memory and mine are coincident - i had the impression that "router" came from Cisco Systems but that had to be after 1984. If you introduced the term "router" in 1977 that would certainly be early. I stuck with "gateways" for quite a while after the 1977 three network demonstration. v On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history < internet-history at elists.isoc.org> wrote: > Watching Jeopardy from November 17, 2021, I was surprised to see a > "clue" which was a picture of Len Kleinrock standing in front of the > ARPANET IMP which has been preserved at UCLA. It was a clue under the > category "It's a New Machine". > > The host read the clue: > > "In 1969, Leonard Kleinrock and his team sent the first message over the > Internet with the first THIS, which now connects devices like modems." > > None of the contestants knew the answer, although one guessed > "backbone", which isn't a bad guess. So the guest revealed the answer: > > "You're looking at the first router." > > My immediate reaction was "No, you're not! That's an IMP." > > See https://www.youtube.com/watch?v=Lmv1WKMGcDw starting at about 1:29 > > That's not quite like I remember it. Ginny Strazisar built the first > switching device for the Internet to connect the ARPANET to the Packet > Radio net, circa 1977. To me that was the genesis of the "Internet" - > INTERconnecting of NETworks of a variety of types, using TCP/IP to glue > it all together. But millions of people just learned otherwise. > > A historical tidbit -- Back in 1977 Ginny's system was called a > "gateway", but later was renamed a "router". It's possible that I did > the renaming. At BBN we were selling lots of packet switches, and > sometimes customers asked for ideas on how to use their LANs in the > network. Our sales people would tell them about the research > activities, and the role of gateways, TCP/IP , and the Internet. But in > many customers' minds that term "gateway" immediately set off alarm > bells, because they had prior bad experience with "gateways" in their > IBM networks, and didn't want anything to do with more "gateways". So > I suggested calling them "routers" instead of "gateways", and suddenly > the marketplace became much more willing to listen. > > Enjoy, > Jack Haverty > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From vgcerf at gmail.com Mon Nov 22 13:36:23 2021 From: vgcerf at gmail.com (vinton cerf) Date: Mon, 22 Nov 2021 16:36:23 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <6769fd88-804c-b41f-9878-ff5eeb340e40@dcrocker.net> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <6769fd88-804c-b41f-9878-ff5eeb340e40@dcrocker.net> Message-ID: interestingly 1989 is the arrival of commercial internet (uunet, psinet, cerfnet). v On Mon, Nov 22, 2021 at 4:14 PM Dave Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > On 11/22/2021 12:50 PM, Jack Haverty via Internet-history wrote: > > > > My immediate reaction was "No, you're not! That's an IMP." > > > Welcome to the distinction between a popular perception versus a > professional one. The former lacks nuance, using very coarse metrics. > The latter ought to be more refined, and sometimes is. > > For the world at large, packet switching is really the defining moment. > > For them, the moment is the invention of computer networking, rather > than the invention of linking networks together. (That is, assuming > that are careful enough to avoid confusing Web with Internet...[*]) > > Sometimes, the error is in the later direction. > > Getting even professionals to be careful in talking about email history > is difficult. So it is quite common even in highly technical circles > -- such as a couple of weeks ago for a press release -- for folk to say > that Ray invented email rather than Ray invented networked email. > > d/ > > [*] maybe 20 years ago, taking a Spanish course in Spain, with a class > of much (much) younger folk from all over Europe, the instructor > prompted some discussion in Spanish by asking us about our backgrounds. > I chose to say that I worked on the UCLA networking project, in 1972, > explaining it was the first site on the Internet. One of the very > bright, very young students objected vigorously, saying that the > Internet was invented in 1989.[**] I smiled and tried to explain the > difference but she persisted. The instructor didn't care about the > answer, as long as everyone was talking in Spanish, but this dragged on. > The youngster would not relent. Finally the instructor intervened, > say "Look, you weren't born yet and he was there!" > > [**] I have heard of a name for it, but there should be one that > distinguishes errors that require a lot of knowledge to make. That she > knew of 1989 in the net's history was impressive. My first encounter > with this type of error was while at the University of Delaware, around > 1980, talking to a hotel reservation agent in Toronto. She asked for my > address and when I said Newark, Delaware, she queried "that's a suburb > of Philadelphia, isn't it?" sigh. > -- > Dave Crocker > Brandenburg InternetWorking > bbiw.net > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From bpurvy at gmail.com Mon Nov 22 13:49:49 2021 From: bpurvy at gmail.com (Bob Purvy) Date: Mon, 22 Nov 2021 13:49:49 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: "backbone" wasn't bad, but it definitely shouldn't have been accepted. We don't know what they'd have ruled if the contestant said something correct, like "gateway" or "IMP." *Alex*: Oh, I'm *so* sorry. The answer is "router." On Mon, Nov 22, 2021 at 1:34 PM vinton cerf via Internet-history < internet-history at elists.isoc.org> wrote: > your memory and mine are coincident - i had the impression that "router" > came from Cisco Systems but that had to be after 1984. If you introduced > the term "router" in 1977 that would certainly be early. I stuck with > "gateways" for quite a while after the 1977 three network demonstration. > > v > > > On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history < > internet-history at elists.isoc.org> wrote: > > > Watching Jeopardy from November 17, 2021, I was surprised to see a > > "clue" which was a picture of Len Kleinrock standing in front of the > > ARPANET IMP which has been preserved at UCLA. It was a clue under the > > category "It's a New Machine". > > > > The host read the clue: > > > > "In 1969, Leonard Kleinrock and his team sent the first message over the > > Internet with the first THIS, which now connects devices like modems." > > > > None of the contestants knew the answer, although one guessed > > "backbone", which isn't a bad guess. So the guest revealed the answer: > > > > "You're looking at the first router." > > > > My immediate reaction was "No, you're not! That's an IMP." > > > > See https://www.youtube.com/watch?v=Lmv1WKMGcDw starting at about 1:29 > > > > That's not quite like I remember it. Ginny Strazisar built the first > > switching device for the Internet to connect the ARPANET to the Packet > > Radio net, circa 1977. To me that was the genesis of the "Internet" - > > INTERconnecting of NETworks of a variety of types, using TCP/IP to glue > > it all together. But millions of people just learned otherwise. > > > > A historical tidbit -- Back in 1977 Ginny's system was called a > > "gateway", but later was renamed a "router". It's possible that I did > > the renaming. At BBN we were selling lots of packet switches, and > > sometimes customers asked for ideas on how to use their LANs in the > > network. Our sales people would tell them about the research > > activities, and the role of gateways, TCP/IP , and the Internet. But in > > many customers' minds that term "gateway" immediately set off alarm > > bells, because they had prior bad experience with "gateways" in their > > IBM networks, and didn't want anything to do with more "gateways". So > > I suggested calling them "routers" instead of "gateways", and suddenly > > the marketplace became much more willing to listen. > > > > Enjoy, > > Jack Haverty > > > > > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From casner at acm.org Mon Nov 22 14:01:36 2021 From: casner at acm.org (Stephen Casner) Date: Mon, 22 Nov 2021 14:01:36 -0800 (PST) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: On Mon, 22 Nov 2021, Bob Purvy via Internet-history wrote: > "backbone" wasn't bad, but it definitely shouldn't have been accepted. We > don't know what they'd have ruled if the contestant said something correct, > like "gateway" or "IMP." > > *Alex*: Oh, I'm *so* sorry. The answer is "router." Or more appropriately, the question is "What is a router." -- Steve From bpurvy at gmail.com Mon Nov 22 14:12:43 2021 From: bpurvy at gmail.com (Bob Purvy) Date: Mon, 22 Nov 2021 14:12:43 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: that was a test, and you passed ? On Mon, Nov 22, 2021 at 2:01 PM Stephen Casner wrote: > On Mon, 22 Nov 2021, Bob Purvy via Internet-history wrote: > > > "backbone" wasn't bad, but it definitely shouldn't have been accepted. We > > don't know what they'd have ruled if the contestant said something > correct, > > like "gateway" or "IMP." > > > > *Alex*: Oh, I'm *so* sorry. The answer is "router." > > Or more appropriately, the question is "What is a router." > > -- Steve > From dan at lynch.com Mon Nov 22 14:37:46 2021 From: dan at lynch.com (Dan Lynch) Date: Mon, 22 Nov 2021 14:37:46 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: <621A290C-8D6E-48DB-B6E2-4726F532D1AE@lynch.com> And all of them were at Interop 89 in SanJose. Now that?s commercial! Dan Cell 650-776-7313 > On Nov 22, 2021, at 1:36 PM, vinton cerf via Internet-history wrote: > > ?interestingly 1989 is the arrival of commercial internet (uunet, psinet, > cerfnet). > > v > > >> On Mon, Nov 22, 2021 at 4:14 PM Dave Crocker via Internet-history < >> internet-history at elists.isoc.org> wrote: >> >>> On 11/22/2021 12:50 PM, Jack Haverty via Internet-history wrote: >>> >>> My immediate reaction was "No, you're not! That's an IMP." >> >> >> Welcome to the distinction between a popular perception versus a >> professional one. The former lacks nuance, using very coarse metrics. >> The latter ought to be more refined, and sometimes is. >> >> For the world at large, packet switching is really the defining moment. >> >> For them, the moment is the invention of computer networking, rather >> than the invention of linking networks together. (That is, assuming >> that are careful enough to avoid confusing Web with Internet...[*]) >> >> Sometimes, the error is in the later direction. >> >> Getting even professionals to be careful in talking about email history >> is difficult. So it is quite common even in highly technical circles >> -- such as a couple of weeks ago for a press release -- for folk to say >> that Ray invented email rather than Ray invented networked email. >> >> d/ >> >> [*] maybe 20 years ago, taking a Spanish course in Spain, with a class >> of much (much) younger folk from all over Europe, the instructor >> prompted some discussion in Spanish by asking us about our backgrounds. >> I chose to say that I worked on the UCLA networking project, in 1972, >> explaining it was the first site on the Internet. One of the very >> bright, very young students objected vigorously, saying that the >> Internet was invented in 1989.[**] I smiled and tried to explain the >> difference but she persisted. The instructor didn't care about the >> answer, as long as everyone was talking in Spanish, but this dragged on. >> The youngster would not relent. Finally the instructor intervened, >> say "Look, you weren't born yet and he was there!" >> >> [**] I have heard of a name for it, but there should be one that >> distinguishes errors that require a lot of knowledge to make. That she >> knew of 1989 in the net's history was impressive. My first encounter >> with this type of error was while at the University of Delaware, around >> 1980, talking to a hotel reservation agent in Toronto. She asked for my >> address and when I said Newark, Delaware, she queried "that's a suburb >> of Philadelphia, isn't it?" sigh. >> -- >> Dave Crocker >> Brandenburg InternetWorking >> bbiw.net >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From jnc at mercury.lcs.mit.edu Mon Nov 22 15:57:19 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Mon, 22 Nov 2021 18:57:19 -0500 (EST) Subject: [ih] "The First Router" on Jeopardy Message-ID: <20211122235719.86E7918C073@mercury.lcs.mit.edu> > From: Jack Haverty > Back in 1977 Ginny's system was called a "gateway", but later was > renamed a "router". ... in many customers' minds that term "gateway" > immediately set off alarm bells, because they had prior bad experience > with "gateways" in their IBM networks, and didn't want anything to do > with more "gateways". I have a very definite memory (although not as detailed as would be optimal) that the IETF community became disenchanted with the term 'gateway' for 'internetworking level packet switch' (the PUP people had earlier called those 'gateways' too - although 'media gateway' was their verbose, specific term - they also had 'protocol gateways'), because all sorts of people started to pop up with boxes they called 'gateways', but which performed a wide range of different functions (usually application level - many of them for email), so one had no idea what someone _meant_ when they said 'and in here we have a gateway'. So we decided we needed a new term; and 'router' (I don't recall who suggested it, alas - too bad we don't have the internet working group amail archives - or maybe we'd transitioned to the IETF mailing list by then) was short and snappy, and everyone liked it, so it was adapted by acclamation. It should be possible to track this by looking at early RFC's, which I'll do in a second. Noel From galmes at tamu.edu Mon Nov 22 16:11:00 2021 From: galmes at tamu.edu (Guy Almes) Date: Mon, 22 Nov 2021 19:11:00 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <20211122235719.86E7918C073@mercury.lcs.mit.edu> References: <20211122235719.86E7918C073@mercury.lcs.mit.edu> Message-ID: <1ab07296-c32e-4ca9-5f88-db17601e25d3@tamu.edu> Noel, That's consistent with my memory. Recall that Cisco's first big product was the AGS, where the G stood for Gateway. I remember the word 'gateway' being used to refer to packet forwarders at various levels in the protocol stack. Eventually, roughly 1987-88, people started to be careful to use 'router' to refer to a level-3 gateway and 'bridge' to a level-2 gateway. And there were application-level gateways such as the various email gateways, including the wonderful box that JANET had in London that reversed the DNS names. A later language shift came a few years later, when the word 'bridge' became old-fashioned and people (or was this just marketing?) began to use the word 'switch' to refer to a high-performance bridge. And, later, the word 'switch' was used also for level-3 packet forwarding. I do not recall where that came from, but did notice it. -- Guy On 11/22/21 6:57 PM, Noel Chiappa via Internet-history wrote: > > > > From: Jack Haverty > > > Back in 1977 Ginny's system was called a "gateway", but later was > > renamed a "router". ... in many customers' minds that term "gateway" > > immediately set off alarm bells, because they had prior bad experience > > with "gateways" in their IBM networks, and didn't want anything to do > > with more "gateways". > > I have a very definite memory (although not as detailed as would be optimal) > that the IETF community became disenchanted with the term 'gateway' for > 'internetworking level packet switch' (the PUP people had earlier called > those 'gateways' too - although 'media gateway' was their verbose, specific > term - they also had 'protocol gateways'), because all sorts of people > started to pop up with boxes they called 'gateways', but which performed a > wide range of different functions (usually application level - many of them > for email), so one had no idea what someone _meant_ when they said 'and in > here we have a gateway'. > > So we decided we needed a new term; and 'router' (I don't recall who > suggested it, alas - too bad we don't have the internet working group amail > archives - or maybe we'd transitioned to the IETF mailing list by then) was > short and snappy, and everyone liked it, so it was adapted by acclamation. > > It should be possible to track this by looking at early RFC's, which I'll > do in a second. > > Noel > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://urldefense.com/v3/__https://elists.isoc.org/mailman/listinfo/internet-history__;!!KwNVnqRv!Uxi3LNpa2KjbwjmZEgLsR86DcNVrzLkUqlKh2AjnWo-eVLALEWlrMl7f4VS5GA$ > From dhc at dcrocker.net Mon Nov 22 16:22:50 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Mon, 22 Nov 2021 16:22:50 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <1ab07296-c32e-4ca9-5f88-db17601e25d3@tamu.edu> References: <20211122235719.86E7918C073@mercury.lcs.mit.edu> <1ab07296-c32e-4ca9-5f88-db17601e25d3@tamu.edu> Message-ID: On 11/22/2021 4:11 PM, Guy Almes via Internet-history wrote: > ? Recall that Cisco's first big product was the AGS, where the G stood > for Gateway. Around 1986 (maybe 1987), at Ungermann-Bass we did a basic IP relaying product and my recollection is that the convention, by then, was the term 'router'. > ? I remember the word 'gateway' being used to refer to packet > forwarders at various levels in the protocol stack. > ? Eventually, roughly 1987-88, people started to be careful to use > 'router' to refer to a level-3 gateway and 'bridge' to a level-2 gateway. > ? And there were application-level gateways such as the various email > gateways, including the wonderful box that JANET had in London that > reversed the DNS names. I believe that was a modified version of my MMDF system that had been... gatewaying email between the Arpanet and telephone-based systems for the Army and then CSNet, since about 1980. As 'gateway' stopped being the term for simple s/f functionality, it came to be used to describe a package that translates between heterogeneous systems. That is, systems that have similarities but are not interoperable. At a minimum, there are differences in formats, but more typically differences in semantics. The differences in semantics turn out to mean that the job of a gateway is to lose information, but only as little as is workable. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From tony.li at tony.li Mon Nov 22 16:53:01 2021 From: tony.li at tony.li (Tony Li) Date: Mon, 22 Nov 2021 16:53:01 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <1ab07296-c32e-4ca9-5f88-db17601e25d3@tamu.edu> References: <20211122235719.86E7918C073@mercury.lcs.mit.edu> <1ab07296-c32e-4ca9-5f88-db17601e25d3@tamu.edu> Message-ID: > On Nov 22, 2021, at 4:11 PM, Guy Almes via Internet-history wrote: > > A later language shift came a few years later, when the word 'bridge' became old-fashioned and people (or was this just marketing?) began to use the word 'switch' to refer to a high-performance bridge. And, later, the word 'switch' was used also for level-3 packet forwarding. > I do not recall where that came from, but did notice it. The word ?switch? was introduced by Kalpana as a marketing term for their ASIC implementation of a bridge. They went that direction because bridging had become a four letter word and they needed to reframe the market. They were successful at that to the point where router vendors started touting ?level-3 switches? in an attempt to further confuse the market. It was too late, the damage was done. T From jack at 3kitty.org Mon Nov 22 18:26:07 2021 From: jack at 3kitty.org (Jack Haverty) Date: Mon, 22 Nov 2021 18:26:07 -0800 Subject: [ih] Interop historical material? In-Reply-To: <621A290C-8D6E-48DB-B6E2-4726F532D1AE@lynch.com> References: <621A290C-8D6E-48DB-B6E2-4726F532D1AE@lynch.com> Message-ID: Speaking of Interop.... Has any of the material from the Interops been preserved??? There were administrative pieces, like lists of vendors, floor maps, titles of sessions, speakers' presentations, attendee lists, and such as well as mountains of presentation slides, papers, vendor handouts, and promotional material. I still have an "Advanced Computing Environments" tile, from before it was called interop, and one of the "Eyeball" giveaways we handed out when we did the "Maze" game on the live show net.? ?? Also I even have a few audiotapes of sessions I gave, although I no longer have any device that can play them. All of this kind of stuff might be historically interesting to someone and useful as a record of what happened when, especially in the "commercial" world rather than the "research" world captured in things like RFCs. Jack On 11/22/21 2:37 PM, Dan Lynch via Internet-history wrote: > And all of them were at Interop 89 in SanJose. Now that?s commercial! > > Dan > > Cell 650-776-7313 > >> On Nov 22, 2021, at 1:36 PM, vinton cerf via Internet-history wrote: >> >> ?interestingly 1989 is the arrival of commercial internet (uunet, psinet, >> cerfnet). >> >> v >> >> >>> On Mon, Nov 22, 2021 at 4:14 PM Dave Crocker via Internet-history < >>> internet-history at elists.isoc.org> wrote: >>> >>>> On 11/22/2021 12:50 PM, Jack Haverty via Internet-history wrote: >>>> >>>> My immediate reaction was "No, you're not! That's an IMP." >>> >>> Welcome to the distinction between a popular perception versus a >>> professional one. The former lacks nuance, using very coarse metrics. >>> The latter ought to be more refined, and sometimes is. >>> >>> For the world at large, packet switching is really the defining moment. >>> >>> For them, the moment is the invention of computer networking, rather >>> than the invention of linking networks together. (That is, assuming >>> that are careful enough to avoid confusing Web with Internet...[*]) >>> >>> Sometimes, the error is in the later direction. >>> >>> Getting even professionals to be careful in talking about email history >>> is difficult. So it is quite common even in highly technical circles >>> -- such as a couple of weeks ago for a press release -- for folk to say >>> that Ray invented email rather than Ray invented networked email. >>> >>> d/ >>> >>> [*] maybe 20 years ago, taking a Spanish course in Spain, with a class >>> of much (much) younger folk from all over Europe, the instructor >>> prompted some discussion in Spanish by asking us about our backgrounds. >>> I chose to say that I worked on the UCLA networking project, in 1972, >>> explaining it was the first site on the Internet. One of the very >>> bright, very young students objected vigorously, saying that the >>> Internet was invented in 1989.[**] I smiled and tried to explain the >>> difference but she persisted. The instructor didn't care about the >>> answer, as long as everyone was talking in Spanish, but this dragged on. >>> The youngster would not relent. Finally the instructor intervened, >>> say "Look, you weren't born yet and he was there!" >>> >>> [**] I have heard of a name for it, but there should be one that >>> distinguishes errors that require a lot of knowledge to make. That she >>> knew of 1989 in the net's history was impressive. My first encounter >>> with this type of error was while at the University of Delaware, around >>> 1980, talking to a hotel reservation agent in Toronto. She asked for my >>> address and when I said Newark, Delaware, she queried "that's a suburb >>> of Philadelphia, isn't it?" sigh. >>> -- >>> Dave Crocker >>> Brandenburg InternetWorking >>> bbiw.net >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history >>> >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history From jmamodio at gmail.com Tue Nov 23 06:34:46 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Tue, 23 Nov 2021 08:34:46 -0600 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: BBN Report #1783 (1969) includes the addition of the "Routing Algorithm" but it does not define the word "router" but the description implies that the IMP with this algorithm acts as a router. Trying to find a copy of Report #1822 .... Cheers Jorge On Mon, Nov 22, 2021 at 2:50 PM Jack Haverty via Internet-history < internet-history at elists.isoc.org> wrote: > Watching Jeopardy from November 17, 2021, I was surprised to see a > "clue" which was a picture of Len Kleinrock standing in front of the > ARPANET IMP which has been preserved at UCLA. It was a clue under the > category "It's a New Machine". > > The host read the clue: > > "In 1969, Leonard Kleinrock and his team sent the first message over the > Internet with the first THIS, which now connects devices like modems." > > None of the contestants knew the answer, although one guessed > "backbone", which isn't a bad guess. So the guest revealed the answer: > > "You're looking at the first router." > > My immediate reaction was "No, you're not! That's an IMP." > > See https://www.youtube.com/watch?v=Lmv1WKMGcDw starting at about 1:29 > > That's not quite like I remember it. Ginny Strazisar built the first > switching device for the Internet to connect the ARPANET to the Packet > Radio net, circa 1977. To me that was the genesis of the "Internet" - > INTERconnecting of NETworks of a variety of types, using TCP/IP to glue > it all together. But millions of people just learned otherwise. > > A historical tidbit -- Back in 1977 Ginny's system was called a > "gateway", but later was renamed a "router". It's possible that I did > the renaming. At BBN we were selling lots of packet switches, and > sometimes customers asked for ideas on how to use their LANs in the > network. Our sales people would tell them about the research > activities, and the role of gateways, TCP/IP , and the Internet. But in > many customers' minds that term "gateway" immediately set off alarm > bells, because they had prior bad experience with "gateways" in their > IBM networks, and didn't want anything to do with more "gateways". So > I suggested calling them "routers" instead of "gateways", and suddenly > the marketplace became much more willing to listen. > > Enjoy, > Jack Haverty > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From jmamodio at gmail.com Tue Nov 23 06:41:14 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Tue, 23 Nov 2021 08:41:14 -0600 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: There is another mention of the "Routing Algorithm" on the 1972 memo from Crowther, Walden and Mimno, to Frank Heart/IMP Guys, describing the new routing algorithm. -J On Tue, Nov 23, 2021 at 8:34 AM Jorge Amodio wrote: > > BBN Report #1783 (1969) includes the addition of the "Routing Algorithm" > but it does not define the word "router" but the description implies that > the IMP with this algorithm acts as a router. > > Trying to find a copy of Report #1822 .... > > Cheers > Jorge > > > On Mon, Nov 22, 2021 at 2:50 PM Jack Haverty via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Watching Jeopardy from November 17, 2021, I was surprised to see a >> "clue" which was a picture of Len Kleinrock standing in front of the >> ARPANET IMP which has been preserved at UCLA. It was a clue under the >> category "It's a New Machine". >> >> The host read the clue: >> >> "In 1969, Leonard Kleinrock and his team sent the first message over the >> Internet with the first THIS, which now connects devices like modems." >> >> None of the contestants knew the answer, although one guessed >> "backbone", which isn't a bad guess. So the guest revealed the answer: >> >> "You're looking at the first router." >> >> My immediate reaction was "No, you're not! That's an IMP." >> >> See https://www.youtube.com/watch?v=Lmv1WKMGcDw starting at about 1:29 >> >> That's not quite like I remember it. Ginny Strazisar built the first >> switching device for the Internet to connect the ARPANET to the Packet >> Radio net, circa 1977. To me that was the genesis of the "Internet" - >> INTERconnecting of NETworks of a variety of types, using TCP/IP to glue >> it all together. But millions of people just learned otherwise. >> >> A historical tidbit -- Back in 1977 Ginny's system was called a >> "gateway", but later was renamed a "router". It's possible that I did >> the renaming. At BBN we were selling lots of packet switches, and >> sometimes customers asked for ideas on how to use their LANs in the >> network. Our sales people would tell them about the research >> activities, and the role of gateways, TCP/IP , and the Internet. But in >> many customers' minds that term "gateway" immediately set off alarm >> bells, because they had prior bad experience with "gateways" in their >> IBM networks, and didn't want anything to do with more "gateways". So >> I suggested calling them "routers" instead of "gateways", and suddenly >> the marketplace became much more willing to listen. >> >> Enjoy, >> Jack Haverty >> >> >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > From vgcerf at gmail.com Tue Nov 23 06:45:33 2021 From: vgcerf at gmail.com (vinton cerf) Date: Tue, 23 Nov 2021 09:45:33 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: yes, there is no dispute that the IMP did routing of packets. But except for Jack's comments, I have never referred to the IMP as a "router," which term I associated with various Internet packet switches. v On Tue, Nov 23, 2021 at 9:42 AM Jorge Amodio via Internet-history < internet-history at elists.isoc.org> wrote: > There is another mention of the "Routing Algorithm" on the 1972 memo from > Crowther, Walden and Mimno, to Frank Heart/IMP Guys, describing the new > routing algorithm. > > -J > > On Tue, Nov 23, 2021 at 8:34 AM Jorge Amodio wrote: > > > > > BBN Report #1783 (1969) includes the addition of the "Routing Algorithm" > > but it does not define the word "router" but the description implies that > > the IMP with this algorithm acts as a router. > > > > Trying to find a copy of Report #1822 .... > > > > Cheers > > Jorge > > > > > > On Mon, Nov 22, 2021 at 2:50 PM Jack Haverty via Internet-history < > > internet-history at elists.isoc.org> wrote: > > > >> Watching Jeopardy from November 17, 2021, I was surprised to see a > >> "clue" which was a picture of Len Kleinrock standing in front of the > >> ARPANET IMP which has been preserved at UCLA. It was a clue under the > >> category "It's a New Machine". > >> > >> The host read the clue: > >> > >> "In 1969, Leonard Kleinrock and his team sent the first message over the > >> Internet with the first THIS, which now connects devices like modems." > >> > >> None of the contestants knew the answer, although one guessed > >> "backbone", which isn't a bad guess. So the guest revealed the answer: > >> > >> "You're looking at the first router." > >> > >> My immediate reaction was "No, you're not! That's an IMP." > >> > >> See https://www.youtube.com/watch?v=Lmv1WKMGcDw starting at about 1:29 > >> > >> That's not quite like I remember it. Ginny Strazisar built the first > >> switching device for the Internet to connect the ARPANET to the Packet > >> Radio net, circa 1977. To me that was the genesis of the "Internet" - > >> INTERconnecting of NETworks of a variety of types, using TCP/IP to glue > >> it all together. But millions of people just learned otherwise. > >> > >> A historical tidbit -- Back in 1977 Ginny's system was called a > >> "gateway", but later was renamed a "router". It's possible that I did > >> the renaming. At BBN we were selling lots of packet switches, and > >> sometimes customers asked for ideas on how to use their LANs in the > >> network. Our sales people would tell them about the research > >> activities, and the role of gateways, TCP/IP , and the Internet. But in > >> many customers' minds that term "gateway" immediately set off alarm > >> bells, because they had prior bad experience with "gateways" in their > >> IBM networks, and didn't want anything to do with more "gateways". So > >> I suggested calling them "routers" instead of "gateways", and suddenly > >> the marketplace became much more willing to listen. > >> > >> Enjoy, > >> Jack Haverty > >> > >> > >> -- > >> Internet-history mailing list > >> Internet-history at elists.isoc.org > >> https://elists.isoc.org/mailman/listinfo/internet-history > >> > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From steve at shinkuro.com Tue Nov 23 06:59:36 2021 From: steve at shinkuro.com (Steve Crocker) Date: Tue, 23 Nov 2021 09:59:36 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: I have no problem referring to the Arpanet IMPs as routers. They weren't *called* routers at the time, but their function was indeed to route packets. So, when explaining things to a modern audience, I say "these boxes were routers -- we called them IMPs -- and it was a key breakthrough in the design process to decide to put the routing function in separate machines." It's of some historical interest to trace the introduction and use of various terms, but, in my view, that's a separate and distinct concern from presenting an understandable description of the function. That said, I'll acknowledge that a bridge, which simply flooded the channels with whatever came in, doesn't quite rise to the level of being a router. I'm not sure whether a switch deserves to be classified as a router, and I'd prefer not to debate it. Steve From jmamodio at gmail.com Tue Nov 23 07:09:37 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Tue, 23 Nov 2021 09:09:37 -0600 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: I guess it could be interesting to find out when the word router was introduced as a "device" Can't find yet anything concrete in those old documents, besides the reference to the routing algorithm, there is also the concept of the IMP performing path selection. Several documents describe the IMP doing "routing" but none of them define the IMP as just a "router" -J On Tue, Nov 23, 2021 at 8:45 AM vinton cerf wrote: > yes, there is no dispute that the IMP did routing of packets. But except > for Jack's comments, I have never referred to the IMP as a "router," which > term I associated with various Internet packet switches. > > v > > > On Tue, Nov 23, 2021 at 9:42 AM Jorge Amodio via Internet-history < > internet-history at elists.isoc.org> wrote: > >> There is another mention of the "Routing Algorithm" on the 1972 memo from >> Crowther, Walden and Mimno, to Frank Heart/IMP Guys, describing the new >> routing algorithm. >> >> -J >> >> On Tue, Nov 23, 2021 at 8:34 AM Jorge Amodio wrote: >> >> > >> > BBN Report #1783 (1969) includes the addition of the "Routing Algorithm" >> > but it does not define the word "router" but the description implies >> that >> > the IMP with this algorithm acts as a router. >> > >> > Trying to find a copy of Report #1822 .... >> > >> > Cheers >> > Jorge >> > >> > >> > On Mon, Nov 22, 2021 at 2:50 PM Jack Haverty via Internet-history < >> > internet-history at elists.isoc.org> wrote: >> > >> >> Watching Jeopardy from November 17, 2021, I was surprised to see a >> >> "clue" which was a picture of Len Kleinrock standing in front of the >> >> ARPANET IMP which has been preserved at UCLA. It was a clue under the >> >> category "It's a New Machine". >> >> >> >> The host read the clue: >> >> >> >> "In 1969, Leonard Kleinrock and his team sent the first message over >> the >> >> Internet with the first THIS, which now connects devices like modems." >> >> >> >> None of the contestants knew the answer, although one guessed >> >> "backbone", which isn't a bad guess. So the guest revealed the answer: >> >> >> >> "You're looking at the first router." >> >> >> >> My immediate reaction was "No, you're not! That's an IMP." >> >> >> >> See https://www.youtube.com/watch?v=Lmv1WKMGcDw starting at about 1:29 >> >> >> >> That's not quite like I remember it. Ginny Strazisar built the first >> >> switching device for the Internet to connect the ARPANET to the Packet >> >> Radio net, circa 1977. To me that was the genesis of the "Internet" - >> >> INTERconnecting of NETworks of a variety of types, using TCP/IP to glue >> >> it all together. But millions of people just learned otherwise. >> >> >> >> A historical tidbit -- Back in 1977 Ginny's system was called a >> >> "gateway", but later was renamed a "router". It's possible that I did >> >> the renaming. At BBN we were selling lots of packet switches, and >> >> sometimes customers asked for ideas on how to use their LANs in the >> >> network. Our sales people would tell them about the research >> >> activities, and the role of gateways, TCP/IP , and the Internet. But in >> >> many customers' minds that term "gateway" immediately set off alarm >> >> bells, because they had prior bad experience with "gateways" in their >> >> IBM networks, and didn't want anything to do with more "gateways". So >> >> I suggested calling them "routers" instead of "gateways", and suddenly >> >> the marketplace became much more willing to listen. >> >> >> >> Enjoy, >> >> Jack Haverty >> >> >> >> >> >> -- >> >> Internet-history mailing list >> >> Internet-history at elists.isoc.org >> >> https://elists.isoc.org/mailman/listinfo/internet-history >> >> >> > >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> > From dhc at dcrocker.net Tue Nov 23 07:18:31 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Tue, 23 Nov 2021 07:18:31 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: <7d9aee50-e092-e88a-dbe4-04e3248df1f8@dcrocker.net> On 11/23/2021 6:59 AM, Steve Crocker via Internet-history wrote: > That said, I'll acknowledge that a bridge, which simply flooded the > channels with whatever came in, What you've described is a repeater. A bridge is a repeater that learns where hosts are and becomes selective about the packets it forwards. (I assume Radia is on this list.) A repeater does not deal with loops. A bridge does. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From jmamodio at gmail.com Tue Nov 23 08:49:19 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Tue, 23 Nov 2021 10:49:19 -0600 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <7d9aee50-e092-e88a-dbe4-04e3248df1f8@dcrocker.net> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <7d9aee50-e092-e88a-dbe4-04e3248df1f8@dcrocker.net> Message-ID: Bridge vs repeater ... Repeater, copies all "packets" from one segment to another. Bridge, copies all "packets" but no broadcast ones from one segment to another. My .02 -J On Tue, Nov 23, 2021 at 9:26 AM Dave Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > On 11/23/2021 6:59 AM, Steve Crocker via Internet-history wrote: > > That said, I'll acknowledge that a bridge, which simply flooded the > > channels with whatever came in, > > > What you've described is a repeater. > > A bridge is a repeater that learns where hosts are and becomes selective > about the packets it forwards. (I assume Radia is on this list.) > > A repeater does not deal with loops. A bridge does. > > d/ > > -- > Dave Crocker > Brandenburg InternetWorking > bbiw.net > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From clemc at ccc.com Tue Nov 23 09:25:28 2021 From: clemc at ccc.com (Clem Cole) Date: Tue, 23 Nov 2021 12:25:28 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <7d9aee50-e092-e88a-dbe4-04e3248df1f8@dcrocker.net> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <7d9aee50-e092-e88a-dbe4-04e3248df1f8@dcrocker.net> Message-ID: On Tue, Nov 23, 2021 at 10:26 AM Dave Crocker via Internet-history < internet-history at elists.isoc.org> wrote: > On 11/23/2021 6:59 AM, Steve Crocker via Internet-history wrote: > > That said, I'll acknowledge that a bridge, which simply flooded the > > channels with whatever came in, > > > What you've described is a repeater. > > A bridge is a repeater that learns where hosts are and becomes selective > about the packets it forwards. (I assume Radia is on this list.) > > A repeater does not deal with loops. A bridge does. > FWIW: I describe the difference to students as an analog process vs. a digital one. The repeater function (like devices used in radio) listen at a specific frequency and other than the time delay inherent within, what goes in, goes out, but the signal strength as been improved (*i.e.* the S/N is corrected), but it's the same information good or bad on both sides of the repeater. Which means that, if the packet is ill formed coming to the circuit, it will be repeated ill formed going out of the circuit. As you point out - thus a repeater can create loops. A bridge (and a router) has (have) some amount of logic inside of it and at least stores and examines some of the bits before copying the information (if not the entire packet before forwarding them). Moreover the information in does not have to be exactly the same what is out. As was said, some filtering of some packets such as broadcast packets may be tossed. But ill formed packets are not forwarded. Plus, the src address of the packet is not necessarily the same for a packet going in as the source going out of a bridge. Which also means that the CRC must be recalculated if any of the information is changed from input to output. Obviously because a new packet is created, the bridge will also improve the S/N in the electrical signal, but the key is that it may have changed the information content as it did it and not all of the information is passed through it and as you importantly point out, because it can filter, it can check for loops and remove packets that are causing same. Finally, at a higher level yet is a router which adds more logic in that it inspects deeper into the packet and understands more about the contents other than the raw bit level of a bridge [ src/dest/len/crc *vs.* actual protocol knowledge ] and may filter and/or modify the contents of the data within the packet itself. The types of filtering and type of packet modifications are even more robust than that of the bridge. That is to say, more data tends to get filtered and thrown away and /or modified as appropriate. This of course is why some of the early products like what 3Com produced were called 'BROUTERS' that did both of the lower (bridge) higher (router) level functions in the same box, since some of the protocols it might not know how to route, and for those packets, it acts like a bridge -- IIRC it bridged 'NetBIOS' packets and could route Netware and IP. From steve at shinkuro.com Tue Nov 23 09:31:14 2021 From: steve at shinkuro.com (Steve Crocker) Date: Tue, 23 Nov 2021 12:31:14 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: <40F5715E-D0CF-43D2-93EE-E9EB39688A90@shinkuro.com> Clem, Interesting set of distinctions. For me, a key attribute of a router is that it chooses which path to send the packet, ie, it *routes* packets. Steve Sent from my iPhone > On Nov 23, 2021, at 12:25 PM, Clem Cole wrote: > > ? > > >> On Tue, Nov 23, 2021 at 10:26 AM Dave Crocker via Internet-history wrote: >> On 11/23/2021 6:59 AM, Steve Crocker via Internet-history wrote: >> > That said, I'll acknowledge that a bridge, which simply flooded the >> > channels with whatever came in, >> >> >> What you've described is a repeater. >> >> A bridge is a repeater that learns where hosts are and becomes selective >> about the packets it forwards. (I assume Radia is on this list.) >> >> A repeater does not deal with loops. A bridge does. > FWIW: I describe the difference to students as an analog process vs. a digital one. The repeater function (like devices used in radio) listen at a specific frequency and other than the time delay inherent within, what goes in, goes out, but the signal strength as been improved (i.e. the S/N is corrected), but it's the same information good or bad on both sides of the repeater. Which means that, if the packet is ill formed coming to the circuit, it will be repeated ill formed going out of the circuit. As you point out - thus a repeater can create loops. > > A bridge (and a router) has (have) some amount of logic inside of it and at least stores and examines some of the bits before copying the information (if not the entire packet before forwarding them). Moreover the information in does not have to be exactly the same what is out. As was said, some filtering of some packets such as broadcast packets may be tossed. But ill formed packets are not forwarded. Plus, the src address of the packet is not necessarily the same for a packet going in as the source going out of a bridge. Which also means that the CRC must be recalculated if any of the information is changed from input to output. Obviously because a new packet is created, the bridge will also improve the S/N in the electrical signal, but the key is that it may have changed the information content as it did it and not all of the information is passed through it and as you importantly point out, because it can filter, it can check for loops and remove packets that are causing same. > > Finally, at a higher level yet is a router which adds more logic in that it inspects deeper into the packet and understands more about the contents other than the raw bit level of a bridge [ src/dest/len/crc vs. actual protocol knowledge ] and may filter and/or modify the contents of the data within the packet itself. The types of filtering and type of packet modifications are even more robust than that of the bridge. That is to say, more data tends to get filtered and thrown away and /or modified as appropriate. > > This of course is why some of the early products like what 3Com produced were called 'BROUTERS' that did both of the lower (bridge) higher (router) level functions in the same box, since some of the protocols it might not know how to route, and for those packets, it acts like a bridge -- IIRC it bridged 'NetBIOS' packets and could route Netware and IP. From clemc at ccc.com Tue Nov 23 09:37:10 2021 From: clemc at ccc.com (Clem Cole) Date: Tue, 23 Nov 2021 12:37:10 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <40F5715E-D0CF-43D2-93EE-E9EB39688A90@shinkuro.com> References: <40F5715E-D0CF-43D2-93EE-E9EB39688A90@shinkuro.com> Message-ID: On Tue, Nov 23, 2021 at 12:31 PM Steve Crocker wrote: > Clem, > > Interesting set of distinctions. For me, a key attribute of a router is > that it chooses which path to send the packet, ie, it *routes* packets. > Agreed. Maybe an additional attribute that I am leaving out is that the router may have more than one output and must have protocol knowledge to decide which interface to direct (route) the information. From jack at 3kitty.org Tue Nov 23 09:48:17 2021 From: jack at 3kitty.org (Jack Haverty) Date: Tue, 23 Nov 2021 09:48:17 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: The distinction I see is more about what a device doesn't do, rather than just what it does. The IMP routed packets, but did a lot more.? It detected errors and retransmitted packets as needed to provide a "reliable byte stream" service.?? It did flow control to the host computers.?? It did congestion control.?? Essentially, it had a functional equivalent of TCP embedded in the IMP code.?? There was even a notion of supporting multiple networks (there's a field in the packet header) to provide a functional equivalent to IP,? although AFAIK that multiple-network capability was never implemented.? It had mechanisms to recover from failures.?? It had mechanisms to facilitate software changes without human intervention. A router routed, and that's pretty much all it did.? It didn't maintain order of data, or guarantee delivery, or prevent duplicates from being delivered.?? It didn't recover from errors or congestion, except for an unreliable attempt to inform the sender that something was wrong.?? All of that kind of functionality was relegated to the attached host computers. A Router, especially the early versions, just routed.? An IMP did a lot more. Jack Haverty On 11/23/21 7:09 AM, Jorge Amodio wrote: > > I guess it could be interesting to find out when the word router was > introduced as a "device" > > Can't find yet anything concrete in those old documents, besides the > reference to the routing algorithm, there is also the concept of the > IMP performing path selection. > > Several documents describe the IMP doing "routing" but none of them > define the IMP as just a "router" > > -J > > > On Tue, Nov 23, 2021 at 8:45 AM vinton cerf > wrote: > > yes, there is no dispute that the IMP did routing of packets. But > except for Jack's comments, I have never referred to the IMP as a > "router," which term I associated with various Internet packet > switches. > > v > > > On Tue, Nov 23, 2021 at 9:42 AM Jorge Amodio via Internet-history > > wrote: > > There is another mention of the "Routing Algorithm" on the > 1972 memo from > Crowther, Walden and Mimno, to Frank Heart/IMP Guys, > describing the new > routing algorithm. > > -J > > On Tue, Nov 23, 2021 at 8:34 AM Jorge Amodio > > wrote: > > > > > BBN Report #1783 (1969) includes the addition of the > "Routing Algorithm" > > but it does not define the word "router" but the description > implies that > > the IMP with this algorithm acts as a router. > > > > Trying to find a copy of Report #1822 .... > > > > Cheers > > Jorge > > > > > > On Mon, Nov 22, 2021 at 2:50 PM Jack Haverty via > Internet-history < > > internet-history at elists.isoc.org > > wrote: > > > >> Watching Jeopardy from November 17, 2021, I was surprised > to see a > >> "clue" which was a picture of Len Kleinrock standing in > front of the > >> ARPANET IMP which has been preserved at UCLA.? It was a > clue under the > >> category "It's a New Machine". > >> > >> The host read the clue: > >> > >> "In 1969, Leonard Kleinrock and his team sent the first > message over the > >> Internet with the first THIS, which now connects devices > like modems." > >> > >> None of the contestants knew the answer, although one guessed > >> "backbone", which isn't a bad guess.? So the guest revealed > the answer: > >> > >> "You're looking at the first router." > >> > >> My immediate reaction was "No, you're not! That's an IMP." > >> > >> See https://www.youtube.com/watch?v=Lmv1WKMGcDw > starting at > about 1:29 > >> > >> That's not quite like I remember it.? Ginny Strazisar built > the first > >> switching device for the Internet to connect the ARPANET to > the Packet > >> Radio net, circa 1977.? To me that was the genesis of the > "Internet" - > >> INTERconnecting of NETworks of a variety of types, using > TCP/IP to glue > >> it all together.? ?But millions of people just learned > otherwise. > >> > >> A historical tidbit -- Back in 1977 Ginny's system was called a > >> "gateway", but later was renamed a "router". ?It's possible > that I did > >> the renaming.? At BBN we were selling lots of packet > switches, and > >> sometimes customers asked for ideas on how to use their > LANs in the > >> network.? Our sales people would tell them about the research > >> activities, and the role of gateways, TCP/IP , and the > Internet. But in > >> many customers' minds that term "gateway" immediately set > off alarm > >> bells, because they had prior bad experience with > "gateways" in their > >> IBM networks, and didn't want anything to do with more > "gateways".? ?So > >> I suggested calling them "routers" instead of "gateways", > and suddenly > >> the marketplace became much more willing to listen. > >> > >> Enjoy, > >> Jack Haverty > >> > >> > >> -- > >> Internet-history mailing list > >> Internet-history at elists.isoc.org > > >> https://elists.isoc.org/mailman/listinfo/internet-history > > >> > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > From steve at shinkuro.com Tue Nov 23 09:51:21 2021 From: steve at shinkuro.com (Steve Crocker) Date: Tue, 23 Nov 2021 12:51:21 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: The words ?protocol knowledge? suggest to me the routing decision depends on which protocol the packet is part of. That goes a step further than I think is necessary. Generally, all of the data packets are viewed as belonging to a single protocol. In the case of the Internet, the protocol is IP. For example, I wouldn?t expect TCP packets to be routed differently from UDP packets. This implies the routing algorithm is tied to a specific protocol layer. Or to put this another way, the IP layer includes the IP protocol itself and the routing algorithm(s) and protocols at that layer. Steve Sent from my iPhone > On Nov 23, 2021, at 12:37 PM, Clem Cole wrote: > > ? > > >> On Tue, Nov 23, 2021 at 12:31 PM Steve Crocker wrote: >> Clem, >> >> Interesting set of distinctions. For me, a key attribute of a router is that it chooses which path to send the packet, ie, it *routes* packets. > Agreed. Maybe an additional attribute that I am leaving out is that the router may have more than one output and must have protocol knowledge to decide which interface to direct (route) the information. From jack at 3kitty.org Tue Nov 23 10:17:15 2021 From: jack at 3kitty.org (Jack Haverty) Date: Tue, 23 Nov 2021 10:17:15 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: On 11/23/21 9:51 AM, Steve Crocker via Internet-history wrote: > I wouldn?t expect TCP packets to be routed differently from UDP packets. But we *did* expect packets to be routed differently, even though they were all IP packets. That was our purpose in putting the Type-Of-Service and Time-To-Live fields in the IP header.? Some packets required more reliable delivery but time didn't matter.? A good example is during transfer of a large file by FTP.? Other packets required fast delivery, since if they arrived too late they weren't useful.? Packet voice is an example. By setting the TTL and TOS fields appropriately, a host computer could advise the IP backbone of routers of what kind of behavior was appropriate for that packet.? A packet which is part of a large file transfer might be routed over a satellite link, since time wasn't critical.?? That role was envisioned for the "Wideband Net". However, a packet which is part of a conversational voice stream might be routed over only terrestrial nets to minimize delay, and if a router along the way determined that it would take too long to get to a destination (TTL would run down), that router could simply discard it. This would require a lot of mechanism in the routers, including multiple, overlapping, and competing routing algorithms, one for each Type Of Service.?? At the time (1980ish) the router hardware didn't have the capability, and we hadn't figured out how to implement such functionality anyway.?? But The Plan was to have an Internet which would support multiple types of service over the available mix of networks, and route packets quite differently even when they were all IP. Then the Internet went commercial in the mid 80s, and it seems that The Plan was lost. AFAIK, such capabilities never got implemented in IP, although something similar must have been done when "multi-protocol routers" appeared in the late 80s.? They could simultaneously run several overlapping "internets" of different protocols.?? When I was involved in operating Oracle's corporate intranet in the early 90s, we had IP, Netware, Appletalk, OSI, and probably a few others, all running simultaneously.?? I don't recommend it; it was a nightmare to operate. Jack Haverty From clemc at ccc.com Tue Nov 23 10:18:45 2021 From: clemc at ccc.com (Clem Cole) Date: Tue, 23 Nov 2021 13:18:45 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: On Tue, Nov 23, 2021 at 12:51 PM Steve Crocker wrote: > The words ?protocol knowledge? suggest to me the routing decision depends > on which protocol the packet is part of. > But this is what many of the early commercial routers did in fact. IP had not yet 'won' and a lot of site had multiple protocols running on their LANs. Some were routable, some were not. But mixed protocols was probably more the norm in commercials and I bet University circles than not. As I said, I know an early 3Com brouter but my memory is there were others. I think DEC made one or two. Noel can tell us what Proeton did. Cisco certainly was multiprotocol at some point. FWIW: we had to work with a couple of them at LCC that knew how to speak a number of protocols. When we took over the Pathworks development for DEC, we had to set up a lab in Burlington that tested of bunch of that stuff and ensure that Pathworks would not do bad things in multiprotocol environments. That said, IIRC even IP had different rules for routing so some type of inspection was needed for it too, but I think that is beside the point. From jack at 3kitty.org Tue Nov 23 10:39:20 2021 From: jack at 3kitty.org (Jack Haverty) Date: Tue, 23 Nov 2021 10:39:20 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> Sorry, I wasn't clear -- the incident I recalled about the sales people and "gateway" versus "router" was probably a few years later - early 80s.? We always called them "gateways" until then.? Jack On 11/22/21 1:34 PM, vinton cerf wrote: > your memory and mine are coincident?- i had the impression that > "router" came from Cisco Systems but that had to be after 1984. If you > introduced the term "router" in 1977 that would certainly be early. I > stuck with "gateways" for quite a while after the 1977 three network > demonstration. > > v > > > On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history > > wrote: > > Watching Jeopardy from November 17, 2021, I was surprised to see a > "clue" which was a picture of Len Kleinrock standing in front of the > ARPANET IMP which has been preserved at UCLA.? It was a clue under > the > category "It's a New Machine". > > The host read the clue: > > "In 1969, Leonard Kleinrock and his team sent the first message > over the > Internet with the first THIS, which now connects devices like modems." > > None of the contestants knew the answer, although one guessed > "backbone", which isn't a bad guess.? So the guest revealed the > answer: > > "You're looking at the first router." > > My immediate reaction was "No, you're not!??? That's an IMP." > > See https://www.youtube.com/watch?v=Lmv1WKMGcDw > starting at about 1:29 > > That's not quite like I remember it.? Ginny Strazisar built the first > switching device for the Internet to connect the ARPANET to the > Packet > Radio net, circa 1977.? To me that was the genesis of the > "Internet" - > INTERconnecting of NETworks of a variety of types, using TCP/IP to > glue > it all together.?? But millions of people just learned otherwise. > > A historical tidbit -- Back in 1977 Ginny's system was called a > "gateway", but later was renamed a "router".?? It's possible that > I did > the renaming.? At BBN we were selling lots of packet switches, and > sometimes customers asked for ideas on how to use their LANs in the > network.? Our sales people would tell them about the research > activities, and the role of gateways, TCP/IP , and the Internet. > But in > many customers' minds that term "gateway"? immediately set off alarm > bells, because they had prior bad experience with "gateways" in their > IBM networks, and didn't want anything to do with more > "gateways".?? So > I suggested calling them "routers" instead of "gateways", and > suddenly > the marketplace became much more willing to listen. > > Enjoy, > Jack Haverty > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > > https://elists.isoc.org/mailman/listinfo/internet-history > > From woody at pch.net Tue Nov 23 12:41:55 2021 From: woody at pch.net (Bill Woodcock) Date: Tue, 23 Nov 2021 21:41:55 +0100 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <20211122235719.86E7918C073@mercury.lcs.mit.edu> <1ab07296-c32e-4ca9-5f88-db17601e25d3@tamu.edu> Message-ID: <98AC1535-5A6D-4E15-8021-CD08B5992341@pch.net> > On Nov 23, 2021, at 1:53 AM, Tony Li via Internet-history wrote: >> On Nov 22, 2021, at 4:11 PM, Guy Almes via Internet-history wrote: >> A later language shift came a few years later, when the word 'bridge' became old-fashioned and people (or was this just marketing?) began to use the word 'switch' to refer to a high-performance bridge. And, later, the word 'switch' was used also for level-3 packet forwarding. >> I do not recall where that came from, but did notice it. > > > The word ?switch? was introduced by Kalpana as a marketing term for their ASIC implementation of a bridge. That was 1990-1991, no? We were doing switching (and calling it switching) in 1988 at Farallon. Started with LocalTalk, and then added Ethernet in 1989, if I have my dates right. I have a very distinct recollection, because I got chewed out for wearing the product T-shirt on the Interop show floor before the release. :-) I believe Barb Tien caught me and threw a hoodie over me and hustled me back to the booth to change out of it. The distinction between bridging and switching that I remember at the time (and still believe) is that bridging was taking packets from one side, regenerating (and reencapsulating) as necessary, and retransmitting out the other side. Or flooding out the other ports, in the case of something with more than two ports. A ?repeater? was a bridge that specifically had two ports, whereas a ?bridge? had two or more ports (but generally more, else you?d have called it a repeater). By contrast, a ?switch? switched packets between ports as necessary, rather than flooding them indiscriminately. ?Layer 3 switching? on the other hand, was marketing talk for ?routing really fast." -Bill -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 833 bytes Desc: Message signed with OpenPGP URL: From tony.li at tony.li Tue Nov 23 14:45:12 2021 From: tony.li at tony.li (Tony Li) Date: Tue, 23 Nov 2021 14:45:12 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: > On Nov 23, 2021, at 10:17 AM, Jack Haverty via Internet-history wrote: > > On 11/23/21 9:51 AM, Steve Crocker via Internet-history wrote: >> I wouldn?t expect TCP packets to be routed differently from UDP packets. > > But we *did* expect packets to be routed differently, even though they were all IP packets. > > That was our purpose in putting the Type-Of-Service and Time-To-Live fields in the IP header. Some packets required more reliable delivery but time didn't matter. A good example is during transfer of a large file by FTP. Other packets required fast delivery, since if they arrived too late they weren't useful. Packet voice is an example. > > By setting the TTL and TOS fields appropriately, a host computer could advise the IP backbone of routers of what kind of behavior was appropriate for that packet. A packet which is part of a large file transfer might be routed over a satellite link, since time wasn't critical. That role was envisioned for the "Wideband Net". However, a packet which is part of a conversational voice stream might be routed over only terrestrial nets to minimize delay, and if a router along the way determined that it would take too long to get to a destination (TTL would run down), that router could simply discard it. > > This would require a lot of mechanism in the routers, including multiple, overlapping, and competing routing algorithms, one for each Type Of Service. At the time (1980ish) the router hardware didn't have the capability, and we hadn't figured out how to implement such functionality anyway. But The Plan was to have an Internet which would support multiple types of service over the available mix of networks, and route packets quite differently even when they were all IP. > > Then the Internet went commercial in the mid 80s, and it seems that The Plan was lost. > > AFAIK, such capabilities never got implemented in IP, although something similar must have been done when "multi-protocol routers" appeared in the late 80s. They could simultaneously run several overlapping "internets" of different protocols. When I was involved in operating Oracle's corporate intranet in the early 90s, we had IP, Netware, Appletalk, OSI, and probably a few others, all running simultaneously. I don't recommend it; it was a nightmare to operate. The Plan was not lost. TOS routing was implemented and fielded by a few vendors. At the time, the market wasn?t ready and it was ignored. Then the crush to integrate with telephony via ATM became the drive and we responded with IntServ. It lost, but ATM then collapsed under its own weight. IntServ went on to beget DiffServ, which is still in use, but is more about queue management than routing algorithms. More complex routing came to the fore in the mid-90?s with RSVP-TE, which is more about routing different traffic types with different constraints than with different algorithms. That has evolved over time and is currently causing a backlash with Segment Routing (and Flex Algo). The jury is still out on whether Segment Routing will flourish. So, The Plan morphed somewhat, but is still widely in use. Different traffic paths for different applications is still possible, but TE is more widely used for traffic placement. Tony From dan at lynch.com Tue Nov 23 14:54:48 2021 From: dan at lynch.com (Dan Lynch) Date: Tue, 23 Nov 2021 14:54:48 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: Steve is right. But he skirts the real world reality that once the engineers give birth to a technology the marketers take over and all is lost with terminology from then on. Sigh? Dan Cell 650-776-7313 > On Nov 23, 2021, at 6:59 AM, Steve Crocker via Internet-history wrote: > > ?I have no problem referring to the Arpanet IMPs as routers. They weren't > *called* routers at the time, but their function was indeed to route > packets. So, when explaining things to a modern audience, I say "these > boxes were routers -- we called them IMPs -- and it was a key > breakthrough in the design process to decide to put the routing function in > separate machines." > > It's of some historical interest to trace the introduction and use of > various terms, but, in my view, that's a separate and distinct concern from > presenting an understandable description of the function. > > That said, I'll acknowledge that a bridge, which simply flooded the > channels with whatever came in, doesn't quite rise to the level of being a > router. I'm not sure whether a switch deserves to be classified as a > router, and I'd prefer not to debate it. > > Steve > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From tte at cs.fau.de Tue Nov 23 15:16:34 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Wed, 24 Nov 2021 00:16:34 +0100 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: On Tue, Nov 23, 2021 at 02:45:12PM -0800, Tony Li via Internet-history wrote: > So, The Plan morphed somewhat, but is still widely in use. Different traffic paths for different applications is still possible, but TE is more widely used for traffic placement. Not sure what the definition of "traffic placement" is, AFAIK, the 99% use-case for any TE solution (especially SR) is what i'd call "SP network capacity optimization". Aka: fitting as much traffic through the SP topology (consisting of a variety of non-equal-cost-path alternatives) as possible. In other words: TE today is about offering LE (Lousy Effort) Internet service for lower capex. Nothing less, nothing more. Obviously, i think there could and should be more. Cheers Toerless > Tony > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From tony.li at tony.li Tue Nov 23 15:52:49 2021 From: tony.li at tony.li (Tony Li) Date: Tue, 23 Nov 2021 15:52:49 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: Toerless, > On Nov 23, 2021, at 3:16 PM, Toerless Eckert wrote: > > Not sure what the definition of "traffic placement" is, AFAIK, the 99% > use-case for any TE solution (especially SR) is what i'd call > "SP network capacity optimization". Aka: fitting as much traffic through > the SP topology (consisting of a variety of non-equal-cost-path alternatives) > as possible. > > In other words: TE today is about offering LE (Lousy Effort) Internet > service for lower capex. Nothing less, nothing more. > > Obviously, i think there could and should be more. That would be nice, but in practice, it hasn?t found a market. Best Effort (or Lousy Effort) typically results in 99.999% of the packets making it to their destination with perfectly acceptable reliability, delay, and jitter. When SP?s have attempted to charge a premium for the remaining 0.001%, they have found that almost all customers were not interested, and those that did pay the premium did not feel that it was worthwhile. The ultra-extreme folks who want the ultimate reliability are not willing to have SPs in the loop in the first place. But back to the point: the mechanisms are all in place for constrained path computation for various different traffic classes and applications if people choose to turn them on. Some do. T From tte at cs.fau.de Tue Nov 23 16:35:44 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Wed, 24 Nov 2021 01:35:44 +0100 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: On Tue, Nov 23, 2021 at 03:52:49PM -0800, Tony Li wrote: > > Obviously, i think there could and should be more. > That would be nice, but in practice, it hasn?t found a market. Best Effort (or Lousy Effort) typically results in 99.999% of the packets making it to their destination with perfectly acceptable reliability, delay, and jitter. When SP?s have attempted to charge a premium for the remaining 0.001%, they have found that almost all customers were not interested, and those that did pay the premium did not feel that it was worthwhile. My experience from enterprises and SP-services-for Enterprises such as L3VPN are somehwat different. QoS was and probably still is a fairly well selling service. Arguably you always need some such service if you need to guarantee services, and for example contribution TV is not using rate adaptive video, and certainly should never want to because at the end of the contribution path you want 100% of the input bandwidth, and not less on a friday evening. Aka: any actual "real-time" traffic will predominantly have trrhoughput, and arguably many will also have latency and even more so reliability requirements. And don't even let me get into the politics of selling better network transport features only as part of much more expensive and often badly worked out application services. IP Multicast for example was for sure never be made available to OTT by access-SPs to their subscribres to protect the SPs own IPTV service offering against competition, and i was often enough involved back when the SPs feared regulation that would have forced them offer such a service. > The ultra-extreme folks who want the ultimate reliability are not willing to have SPs in the loop in the first place. Unless they have a service where on one side you have a larger customer base that you don't want/cant wire up yourself. > But back to the point: the mechanisms are all in place for constrained path computation for various different traffic classes and applications if people choose to turn them on. Some do. We have a bunch of 25 year old wierd building blocks, but SPs have to business concept competency to make something from them. I do also count mostly on challengers building better networks for future services than expecting existing SPs to figure out how to make new money from new opportunities. Cheers Toerless > T > -- --- tte at cs.fau.de From sob at sobco.com Tue Nov 23 16:43:49 2021 From: sob at sobco.com (Scott Bradner) Date: Tue, 23 Nov 2021 19:43:49 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: > On Nov 23, 2021, at 7:35 PM, Toerless Eckert via Internet-history wrote: > And don't even let me get into the politics of selling better network > transport features only as part of much more expensive and often badly > worked out application services. IP Multicast for example was for sure > never be made available to OTT by access-SPs to their subscribres to > protect the SPs own IPTV service offering against competition, and > i was often enough involved back when the SPs feared regulation that > would have forced them offer such a service. MCI offered many to many IP Multicast as a service but did not get many takers (and figuring out the billing was rather tricky - since the packets could come from anywhere Scott From brian.e.carpenter at gmail.com Tue Nov 23 17:54:55 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Wed, 24 Nov 2021 14:54:55 +1300 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: <13dc8e25-5d8e-21be-c23d-4d637f666d06@gmail.com> Jack, On 24-Nov-21 07:17, Jack Haverty via Internet-history wrote: > On 11/23/21 9:51 AM, Steve Crocker via Internet-history wrote: >> I wouldn?t expect TCP packets to be routed differently from UDP packets. > > But we *did* expect packets to be routed differently, even though they > were all IP packets. > > That was our purpose in putting the Type-Of-Service and Time-To-Live > fields in the IP header.? Some packets required more reliable delivery > but time didn't matter.? A good example is during transfer of a large > file by FTP.? Other packets required fast delivery, since if they > arrived too late they weren't useful.? Packet voice is an example. > > By setting the TTL and TOS fields appropriately, a host computer could > advise the IP backbone of routers of what kind of behavior was > appropriate for that packet.? A packet which is part of a large file > transfer might be routed over a satellite link, since time wasn't > critical.?? That role was envisioned for the "Wideband Net". However, a > packet which is part of a conversational voice stream might be routed > over only terrestrial nets to minimize delay, and if a router along the > way determined that it would take too long to get to a destination (TTL > would run down), that router could simply discard it. > > This would require a lot of mechanism in the routers, Exactly, and it was simply not viable economically, compared to "throwing bandwidth at the problem". > including > multiple, overlapping, and competing routing algorithms, one for each > Type Of Service.?? At the time (1980ish) the router hardware didn't have > the capability, and we hadn't figured out how to implement such > functionality anyway.?? But The Plan was to have an Internet which would > support multiple types of service over the available mix of networks, We tried to recuperate that plan around 1998 (RFC2474, RFC2475), because people were very much aware of those 8 lost bits, and we even incorporated the same 8 bits in the IPv6 header. > and route packets quite differently even when they were all IP. s/route/queue/ and that's what we did, but it proved successful only on site-wide networks, compared to throwing bandwidth at the problem on the WAN. > Then the Internet went commercial in the mid 80s, and it seems that The > Plan was lost. The "New IP" gang and the true believers in Deterministic Networking seem to believe that we can do it now, at least in limited domains. It remains to be seen whether that's viable, compared to (guess what) throwing bandwidth at the problem. The problem's always been the same: can we process packets faster than line speed, which is the only way to sidestep queueing theory? If not, add bandwidth and you're done. (Reading both volumes of Kleinrock's queueing theory book was one of best investments of my time that I can recall. I didn't retain much of the algebra, but the basic messages were loud and clear.) Brian > > AFAIK, such capabilities never got implemented in IP, although something > similar must have been done when "multi-protocol routers" appeared in > the late 80s.? They could simultaneously run several overlapping > "internets" of different protocols.?? When I was involved in operating > Oracle's corporate intranet in the early 90s, we had IP, Netware, > Appletalk, OSI, and probably a few others, all running simultaneously. > I don't recommend it; it was a nightmare to operate. > > Jack Haverty > > From jack at 3kitty.org Tue Nov 23 18:14:10 2021 From: jack at 3kitty.org (Jack Haverty) Date: Tue, 23 Nov 2021 18:14:10 -0800 Subject: [ih] Internet as a Set of Services (IAASOS) -- was Re: "The First Router" on Jeopardy In-Reply-To: References: Message-ID: Changed the name, this has morphed well away from Jeopardy. I look at this now from an end-users' perspective, which I think is quite different from that of a network operator. Whatever kind of services are implemented in a backbone, several other things sometimes have to happen for it to be visible to the end-user: 1/ Some way of selecting a particular service must be exposed at the edge of the network, perhaps in a packet header or protocol exchange.? E.g., the TOS field. 2/ The end-user's OS must act appropriately to use the ability to select the service.? E.g., it must set the TOS field. 3/ If an application has to make the decision about which service to select, the OS must expose that choice in it's interface to user programs, e.g., in its API. E.g., the app has to select the appropriate kind of service. 4/ An application that can benefit from the service has to be designed to select the appropriate service when interfacing with the OS, and possibly also present appropriate choices to the end-user. E.g., the App must be aware of and properly use the OS API to select the appropriate service for whatever it is doing. and lastly, 5/ if the end-user is interacting with a remote machine that requires an Internet pathway through several network operators, all of them must implement the service and pass the appropriate information across the boundaries between them.? E.g., EGP/BGP/etc have to be revised to support service parameters, and networks have to use them. Until all of those steps are taken, it's not surprising that there is no market demand, even if some underlying network mechanism is perfectly designed and implemented in routers or such devices. There is simply no benefit visible to the end-users.?? Zero supply means zero demand. I've found it consistently difficult to even determine if a particular situation meets those criteria above. ? Of course I'm just an end-user, so I shouldn't have to see such details. ? I have 60+ devices on my home LAN at the moment.? I wonder how many of them have the capability to select some specific service from "the network", or if they do whether or not the networks I'm using behave any differently.?? Hard to tell, as the end user, much about what's going on inside all those boxes. But I can see the aggregate effects of whatever is going on "under the covers".?? For example, it's common to see video and audio break up, pixellate, and otherwise misbehave, even on mass-market television channels, especially with events involving participants in remote roundtables (think Zoom, Webex, etc.). A couple of years ago, I helped a friend troubleshoot a "gaming" app he was trying to use on an Internet path between the Los Angeles and Reno Nevada areas.? That involved probably 4 or more distinct networks along the way.?? We couldn't see much except what Wireshark et al could see at his end, but the results were interesting.? Over many tests, 100.0% of the packets sent were correctly received. Not a single packet was dropped. But occasionally, traffic would stop for a while, then resume.? All packets would eventually be delivered.?? Some took 30 seconds in transit.? So, 100% of the packets were delivered, but it's hard to tell how many of them were actually useful by the time they got there.? We couldn't see "inside" the app.? A packet containing audio data that should have been sent to the speaker 2 seconds ago wasn't useful.?? In the TOS/TTL design of "The Plan" such packets would have been discarded as soon as some router determined that they couldn't get there in time to be useful.? It was hard to tell if the OS, or the apps running at each end of the connection, were doing anything special to try to select some type of service, or if the networks even made such a choice possible. Regardless of what was going on, the gaming app was essentially unusable.?? That customer was not "ultra extreme".?? He just wanted to play the game.?? He followed the Marketing advice to "upgrade his service" to higher speeds.? Cost more, but didn't help.?? Even switch his local ISP.? No effect. My anecdotal test with the Gamer was just one tiny piece of the Internet system.?? But it seems like that experience is somewhat common.??? Gamers are known for demanding "low latency" and I've even heard of cases where a Gamer refuses to live somewhere because the Internet service is unusable, for Gamers.?? Even at "gigabit speed".?? Even though there is a market for it. When the Internet was just an ARPA research project, "The Plan" was to have it capable of supporting such scenarios, assuming of course that we could figure out how to do so.? Then the Internet took off, became a DoD Standard, and escaped from the lab.?? Research was replaced by commercial and marketplace concerns. My tentative conclusion is that there are now simply too many uncoordinated and/or competing components in today's "internet system", with no one responsible for making the overall system work to meet the users' needs.?? The original ARPA vision and "The Plan" still seems possible, but no one seems to be working on it.?? Maybe it just no longer matters. Jack Haverty On 11/23/21 4:35 PM, Toerless Eckert wrote: > On Tue, Nov 23, 2021 at 03:52:49PM -0800, Tony Li wrote: >>> Obviously, i think there could and should be more. >> That would be nice, but in practice, it hasn?t found a market. Best Effort (or Lousy Effort) typically results in 99.999% of the packets making it to their destination with perfectly acceptable reliability, delay, and jitter. When SP?s have attempted to charge a premium for the remaining 0.001%, they have found that almost all customers were not interested, and those that did pay the premium did not feel that it was worthwhile. > My experience from enterprises and SP-services-for Enterprises such as L3VPN > are somehwat different. QoS was and probably still is a fairly well > selling service. Arguably you always need some such service if you need > to guarantee services, and for example contribution TV is not using > rate adaptive video, and certainly should never want to because at > the end of the contribution path you want 100% of the input bandwidth, > and not less on a friday evening. Aka: any actual "real-time" traffic > will predominantly have trrhoughput, and arguably many will also have > latency and even more so reliability requirements. > > And don't even let me get into the politics of selling better network > transport features only as part of much more expensive and often badly > worked out application services. IP Multicast for example was for sure > never be made available to OTT by access-SPs to their subscribres to > protect the SPs own IPTV service offering against competition, and > i was often enough involved back when the SPs feared regulation that > would have forced them offer such a service. > >> The ultra-extreme folks who want the ultimate reliability are not willing to have SPs in the loop in the first place. > Unless they have a service where on one side you have a larger > customer base that you don't want/cant wire up yourself. > >> But back to the point: the mechanisms are all in place for constrained path computation for various different traffic classes and applications if people choose to turn them on. Some do. > We have a bunch of 25 year old wierd building blocks, but > SPs have to business concept competency to make something from them. > > I do also count mostly on challengers building better networks > for future services than expecting existing SPs to figure out > how to make new money from new opportunities. > > Cheers > Toerless > >> T >> From j at shoch.com Tue Nov 23 18:25:45 2021 From: j at shoch.com (John Shoch) Date: Tue, 23 Nov 2021 18:25:45 -0800 Subject: [ih] "The First Router" on Jeopardy Message-ID: OK, because of Jeopardy, we are here again: Arpanet vs Internet, IMP vs packet switch vs gateway vs router, internetworking vs Internet, etc., etc....... I'm sitting here suffering from complete sensory overload: --I can "see" Vint Cerf, sitting there, smiling calmly, the epitome of restraint and grace in the midst of ongoing confusion. --I can "feel" Bob Metcalfe, gently kicking me under the table, and whispering: "Let it go, John....." --I can "hear" Bob Taylor, spinning in his grave, yelling at me, "Speak up you idiot! You promised me you would speak up!" Moments like this, when I fibrillate because of conflicting inputs, only get me in trouble.....sigh: a. The Jeopardy answer was posed as, "in 1969...sent the first message over the Internet." The answer is poorly framed, and there can be no correct question. That box (designed and built by BBN), no matter what you call it, was not in 1969 the first to send bits over the Internet. b. I wonder how many other topics Jeopardy has mis-attributed. c. At the time, we would have called it a switch or a packet switch; internetworking gateway emerged later, and then morphed into a router. An IMP certainly did routing, but most of us would say that "router" is a shortened version of "internet router." d. Ginny, BBN, and others deserve a tremendous amount of credit for implementing what was then a gateway, now an Internet router (with an uppercase I) for TCP/IP. As acknowledged by the Computer History Museum: https://www.computerhistory.org/revolution/networking/19/375/2071 e. But we often forget that the process was a lot of work, with an evolution over an extended period of time. [One small example, just to remind us of what it was like over 40 years ago, from IEN #25 in 1978: "The gateways currently use a static routing procedure based on routing tables assembled into each gateway. In the near future, we plan to implement a simple gateway routing scheme, which will improve internet performance by providing the capability to route around failed gateways and networks." https://www.rfc-editor.org/ien/ien25.pdf To repeat, 1) it was the best IP/TCP gateway of it's time, 1978, but 2) in order to add a new network you had to reassemble the code for all gateways, and 3) there was not yet dynamic routing around a failed network or gateway......] f. And let's not forget the earlier implementation of an internet gateway (with a lowercase i) by Ed Taft and Dave Boggs [not me], ca. 1974-1975. Also from the CHM: https://www.computerhistory.org/revolution/networking/19/375/2090 Jeopardy research must be hard; so is computer history. John > From sghuter at nsrc.org Tue Nov 23 18:53:18 2021 From: sghuter at nsrc.org (Steven G. Huter) Date: Tue, 23 Nov 2021 18:53:18 -0800 (PST) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: this is awesome! thank you for the excellent summation of the history and the key highlights. steve huter On Tue, 23 Nov 2021, John Shoch via Internet-history wrote: > OK, because of Jeopardy, we are here again: Arpanet vs Internet, IMP vs > packet switch vs gateway vs router, internetworking vs Internet, etc., > etc....... > > I'm sitting here suffering from complete sensory overload: > --I can "see" Vint Cerf, sitting there, smiling calmly, the epitome of > restraint and grace in the midst of ongoing confusion. > --I can "feel" Bob Metcalfe, gently kicking me under the table, and > whispering: "Let it go, John....." > --I can "hear" Bob Taylor, spinning in his grave, yelling at me, "Speak up > you idiot! You promised me you would speak up!" > > Moments like this, when I fibrillate because of conflicting inputs, only > get me in trouble.....sigh: > > a. The Jeopardy answer was posed as, "in 1969...sent the first message > over the Internet." The answer is poorly framed, and there can be no > correct question. That box (designed and built by BBN), no matter what > you call it, was not in 1969 the first to send bits over the Internet. > b. I wonder how many other topics Jeopardy has mis-attributed. > c. At the time, we would have called it a switch or a packet switch; > internetworking gateway emerged later, and then morphed into a router. An > IMP certainly did routing, but most of us would say that "router" is a > shortened version of "internet router." > d. Ginny, BBN, and others deserve a tremendous amount of credit for > implementing what was then a gateway, now an Internet router (with an > uppercase I) for TCP/IP. As acknowledged by the Computer History Museum: > https://www.computerhistory.org/revolution/networking/19/375/2071 > e. But we often forget that the process was a lot of work, with an > evolution over an extended period of time. > [One small example, just to remind us of what it was like over 40 years > ago, from IEN #25 in 1978: > "The gateways currently use a static routing procedure based on routing > tables assembled into each gateway. In the near future, we plan to > implement a simple gateway routing scheme, which will improve internet > performance by providing the capability to route around failed gateways and > networks." https://www.rfc-editor.org/ien/ien25.pdf To repeat, 1) it was > the best IP/TCP gateway of it's time, 1978, but 2) in order to add a new > network you had to reassemble the code for all gateways, and 3) there was > not yet dynamic routing around a failed network or gateway......] > f. And let's not forget the earlier implementation of an internet gateway > (with a lowercase i) by Ed Taft and Dave Boggs [not me], ca. 1974-1975. > Also from the CHM: > https://www.computerhistory.org/revolution/networking/19/375/2090 > > Jeopardy research must be hard; so is computer history. > > John > > >> > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From b_a_denny at yahoo.com Tue Nov 23 21:19:22 2021 From: b_a_denny at yahoo.com (Barbara Denny) Date: Wed, 24 Nov 2021 05:19:22 +0000 (UTC) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> Message-ID: <1800874110.4972359.1637731162424@mail.yahoo.com> I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. barbara On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote: Sorry, I wasn't clear -- the incident I recalled about the sales people and "gateway" versus "router" was probably a few years later - early 80s.? We always called them "gateways" until then.? Jack On 11/22/21 1:34 PM, vinton cerf wrote: > your memory and mine are coincident?- i had the impression that > "router" came from Cisco Systems but that had to be after 1984. If you > introduced the term "router" in 1977 that would certainly be early. I > stuck with "gateways" for quite a while after the 1977 three network > demonstration. > > v > > > On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history > > wrote: > >? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >? ? "clue" which was a picture of Len Kleinrock standing in front of the >? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >? ? the >? ? category "It's a New Machine". > >? ? The host read the clue: > >? ? "In 1969, Leonard Kleinrock and his team sent the first message >? ? over the >? ? Internet with the first THIS, which now connects devices like modems." > >? ? None of the contestants knew the answer, although one guessed >? ? "backbone", which isn't a bad guess.? So the guest revealed the >? ? answer: > >? ? "You're looking at the first router." > >? ? My immediate reaction was "No, you're not!??? That's an IMP." > >? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >? ? starting at about 1:29 > >? ? That's not quite like I remember it.? Ginny Strazisar built the first >? ? switching device for the Internet to connect the ARPANET to the >? ? Packet >? ? Radio net, circa 1977.? To me that was the genesis of the >? ? "Internet" - >? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >? ? glue >? ? it all together.?? But millions of people just learned otherwise. > >? ? A historical tidbit -- Back in 1977 Ginny's system was called a >? ? "gateway", but later was renamed a "router".?? It's possible that >? ? I did >? ? the renaming.? At BBN we were selling lots of packet switches, and >? ? sometimes customers asked for ideas on how to use their LANs in the >? ? network.? Our sales people would tell them about the research >? ? activities, and the role of gateways, TCP/IP , and the Internet. >? ? But in >? ? many customers' minds that term "gateway"? immediately set off alarm >? ? bells, because they had prior bad experience with "gateways" in their >? ? IBM networks, and didn't want anything to do with more >? ? "gateways".?? So >? ? I suggested calling them "routers" instead of "gateways", and >? ? suddenly >? ? the marketplace became much more willing to listen. > >? ? Enjoy, >? ? Jack Haverty > > >? ? -- >? ? Internet-history mailing list >? ? Internet-history at elists.isoc.org >? ? >? ? https://elists.isoc.org/mailman/listinfo/internet-history >? ? > -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From b_a_denny at yahoo.com Wed Nov 24 00:44:42 2021 From: b_a_denny at yahoo.com (Barbara Denny) Date: Wed, 24 Nov 2021 08:44:42 +0000 (UTC) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <1800874110.4972359.1637731162424@mail.yahoo.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> Message-ID: <2123601323.5011921.1637743482602@mail.yahoo.com> I should also mention the RIG (Research Internet Gateway) Program.? This program was started in 1988 and an online? NASA Ames paper from 1994 uses both terms,? gateways and routers. https://www.nas.nasa.gov/assets/pdf/techreports/1994/rnd-94-008.pdf I didn't look for any reports from each contractor: SRI/Cisco, BBN, and GTE/Proteon. barbara On Tuesday, November 23, 2021, 09:23:12 PM PST, Barbara Denny via Internet-history wrote: I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. barbara ? ? On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote:? Sorry, I wasn't clear -- the incident I recalled about the sales people and "gateway" versus "router" was probably a few years later - early 80s.? We always called them "gateways" until then.? Jack On 11/22/21 1:34 PM, vinton cerf wrote: > your memory and mine are coincident?- i had the impression that > "router" came from Cisco Systems but that had to be after 1984. If you > introduced the term "router" in 1977 that would certainly be early. I > stuck with "gateways" for quite a while after the 1977 three network > demonstration. > > v > > > On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history > > wrote: > >? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >? ? "clue" which was a picture of Len Kleinrock standing in front of the >? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >? ? the >? ? category "It's a New Machine". > >? ? The host read the clue: > >? ? "In 1969, Leonard Kleinrock and his team sent the first message >? ? over the >? ? Internet with the first THIS, which now connects devices like modems." > >? ? None of the contestants knew the answer, although one guessed >? ? "backbone", which isn't a bad guess.? So the guest revealed the >? ? answer: > >? ? "You're looking at the first router." > >? ? My immediate reaction was "No, you're not!??? That's an IMP." > >? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >? ? starting at about 1:29 > >? ? That's not quite like I remember it.? Ginny Strazisar built the first >? ? switching device for the Internet to connect the ARPANET to the >? ? Packet >? ? Radio net, circa 1977.? To me that was the genesis of the >? ? "Internet" - >? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >? ? glue >? ? it all together.?? But millions of people just learned otherwise. > >? ? A historical tidbit -- Back in 1977 Ginny's system was called a >? ? "gateway", but later was renamed a "router".?? It's possible that >? ? I did >? ? the renaming.? At BBN we were selling lots of packet switches, and >? ? sometimes customers asked for ideas on how to use their LANs in the >? ? network.? Our sales people would tell them about the research >? ? activities, and the role of gateways, TCP/IP , and the Internet. >? ? But in >? ? many customers' minds that term "gateway"? immediately set off alarm >? ? bells, because they had prior bad experience with "gateways" in their >? ? IBM networks, and didn't want anything to do with more >? ? "gateways".?? So >? ? I suggested calling them "routers" instead of "gateways", and >? ? suddenly >? ? the marketplace became much more willing to listen. > >? ? Enjoy, >? ? Jack Haverty > > >? ? -- >? ? Internet-history mailing list >? ? Internet-history at elists.isoc.org >? ? >? ? https://elists.isoc.org/mailman/listinfo/internet-history >? ? > -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history ? -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From cabo at tzi.org Wed Nov 24 02:54:25 2021 From: cabo at tzi.org (Carsten Bormann) Date: Wed, 24 Nov 2021 11:54:25 +0100 Subject: [ih] The term "router" in RFCs (Re: "The First Router" on Jeopardy) In-Reply-To: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> Message-ID: <3A07A8E7-7696-4CD4-A152-93F322CD755E@tzi.org> On 2021-11-22, at 21:50, Jack Haverty via Internet-history wrote: > > A historical tidbit -- Back in 1977 Ginny's system was called a "gateway", but later was renamed a "router". It's possible that I did the renaming. At BBN we were selling lots of packet switches, and sometimes customers asked for ideas on how to use their LANs in the network. Our sales people would tell them about the research activities, and the role of gateways, TCP/IP , and the Internet. But in many customers' minds that term "gateway" immediately set off alarm bells, because they had prior bad experience with "gateways" in their IBM networks, and didn't want anything to do with more "gateways". So I suggested calling them "routers" instead of "gateways", and suddenly the marketplace became much more willing to listen. A quick look at using the word ?router? in RFCs/IENs: RFC 753 (= IEN 85, March 1979) used the word for a message router, as did its update RFC 759 (= IEN 113, August 1980). RFC 850 (Usenet messages, June 1983) used ?mail router?. IEN 178, April 1981, uses ?router? once to describe an entity that did network route processing; the somewhat casual usage doesn?t necessarily imply a forwarding function being part of the ?router? concept. The first trace I can find of an Internet gateway being called a router in an RFC is RFC 898 (Gateway SIG Meeting Notes, April 1984): This talks about the CMU gateway, including the lines: History - o "Logical-Host" multiplexor (March 81) o Gateway (Oct 82) remote debugger and monitor o Router (Oct 83) - Modular device and protocol support - Stub IP dynamic routing - Local inter-network cable routing. Apparently, ?router? implied more functionality than a simple ?gateway?, and ?Stub IP dynamic routing? and ?local inter-network cable routing" seemed to be parts of that. So this quick look seems to locate early use of the term with today?s meaning in the early 1980s. Gr??e, Carsten From vint at google.com Wed Nov 24 04:04:08 2021 From: vint at google.com (Vint Cerf) Date: Wed, 24 Nov 2021 07:04:08 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: still smiling - thanks John for clarity, as always. v On Tue, Nov 23, 2021 at 9:26 PM John Shoch via Internet-history < internet-history at elists.isoc.org> wrote: > OK, because of Jeopardy, we are here again: Arpanet vs Internet, IMP vs > packet switch vs gateway vs router, internetworking vs Internet, etc., > etc....... > > I'm sitting here suffering from complete sensory overload: > --I can "see" Vint Cerf, sitting there, smiling calmly, the epitome of > restraint and grace in the midst of ongoing confusion. > --I can "feel" Bob Metcalfe, gently kicking me under the table, and > whispering: "Let it go, John....." > --I can "hear" Bob Taylor, spinning in his grave, yelling at me, "Speak up > you idiot! You promised me you would speak up!" > > Moments like this, when I fibrillate because of conflicting inputs, only > get me in trouble.....sigh: > > a. The Jeopardy answer was posed as, "in 1969...sent the first message > over the Internet." The answer is poorly framed, and there can be no > correct question. That box (designed and built by BBN), no matter what > you call it, was not in 1969 the first to send bits over the Internet. > b. I wonder how many other topics Jeopardy has mis-attributed. > c. At the time, we would have called it a switch or a packet switch; > internetworking gateway emerged later, and then morphed into a router. An > IMP certainly did routing, but most of us would say that "router" is a > shortened version of "internet router." > d. Ginny, BBN, and others deserve a tremendous amount of credit for > implementing what was then a gateway, now an Internet router (with an > uppercase I) for TCP/IP. As acknowledged by the Computer History Museum: > https://www.computerhistory.org/revolution/networking/19/375/2071 > e. But we often forget that the process was a lot of work, with an > evolution over an extended period of time. > [One small example, just to remind us of what it was like over 40 years > ago, from IEN #25 in 1978: > "The gateways currently use a static routing procedure based on routing > tables assembled into each gateway. In the near future, we plan to > implement a simple gateway routing scheme, which will improve internet > performance by providing the capability to route around failed gateways and > networks." https://www.rfc-editor.org/ien/ien25.pdf To repeat, 1) it was > the best IP/TCP gateway of it's time, 1978, but 2) in order to add a new > network you had to reassemble the code for all gateways, and 3) there was > not yet dynamic routing around a failed network or gateway......] > f. And let's not forget the earlier implementation of an internet gateway > (with a lowercase i) by Ed Taft and Dave Boggs [not me], ca. 1974-1975. > Also from the CHM: > https://www.computerhistory.org/revolution/networking/19/375/2090 > > Jeopardy research must be hard; so is computer history. > > John > > > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > -- Please send any postal/overnight deliveries to: Vint Cerf 1435 Woodhurst Blvd McLean, VA 22102 703-448-0965 until further notice From jnc at mercury.lcs.mit.edu Wed Nov 24 05:17:07 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Wed, 24 Nov 2021 08:17:07 -0500 (EST) Subject: [ih] "The First Router" on Jeopardy Message-ID: <20211124131707.25E0C18C074@mercury.lcs.mit.edu> > From: Steve Crocker > I have no problem referring to the Arpanet IMPs as routers. ... So, > when explaining things to a modern audience I'm waiting for the news stories about the 'COVID bacterium'. Oh, and about the latest 'atom' found in the LHC. I trust my point is semi-obvious, but if not: I think we do a dis-service to the public when we don't use the correct terms, and urge them to do likewise. ("Any scientist who can't explain to an eight-year old what he is doing is a charlatan." - Dr. Felix Hoenikker) This margin is way too small for a full discussion of this important point, alas. I understand that one might not want to emit, and the listener might not be interested in receiving, a lesson on the ISO model (upgraded to include separate 'network' and 'internetwork' layers - in its original form, it's broken), so one can clearly define why an IMP is not a router. So, one could just say 'it's a packet switch; it's like a router, but with an important difference'. The apparatus that Hahn and Strassmann (with the later assistance of Lise Meitner, who was in Denmark at the time) used to discover fission wasn't a 'reactor', either, although it did split atoms. Noel From amckenzie3 at yahoo.com Wed Nov 24 07:29:59 2021 From: amckenzie3 at yahoo.com (Alex McKenzie) Date: Wed, 24 Nov 2021 15:29:59 +0000 (UTC) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: <315942109.667960.1637767799111@mail.yahoo.com> I've stayed out of the discussion but listened with interest.? In fact, I have no clear recollection of when the terminology switched.? But I suspect it happened around the time the first CIX (Commercial Internet Exchange) agreement was reached.? Before that there was a "backbone" network (first ARPAnet and then NSFnet).? Packets went from your network to the backbone, across the backbone, and then to the destination network.? But after the CIX agreement there became a meaningful choice of routes, and the gateways had to start figuring out what the set of available routes was and choose one.? I understand this is an oversimplification of the internet structure, but I think it is a meaningful one. Cheers,Alex From gregskinner0 at icloud.com Wed Nov 24 09:15:56 2021 From: gregskinner0 at icloud.com (Greg Skinner) Date: Wed, 24 Nov 2021 09:15:56 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: > On Nov 23, 2021, at 6:25 PM, John Shoch wrote: > > OK, because of Jeopardy, we are here again: Arpanet vs Internet, IMP vs > packet switch vs gateway vs router, internetworking vs Internet, etc., > etc....... > > I'm sitting here suffering from complete sensory overload: > --I can "see" Vint Cerf, sitting there, smiling calmly, the epitome of > restraint and grace in the midst of ongoing confusion. > --I can "feel" Bob Metcalfe, gently kicking me under the table, and > whispering: "Let it go, John....." > --I can "hear" Bob Taylor, spinning in his grave, yelling at me, "Speak up > you idiot! You promised me you would speak up!" > > Moments like this, when I fibrillate because of conflicting inputs, only > get me in trouble.....sigh: > > a. The Jeopardy answer was posed as, "in 1969...sent the first message > over the Internet." The answer is poorly framed, and there can be no > correct question. That box (designed and built by BBN), no matter what > you call it, was not in 1969 the first to send bits over the Internet. Arguably, the correct question was in the picture. :) > b. I wonder how many other topics Jeopardy has mis-attributed. In all fairness to the Jeopardy researchers, there are many articles, videos, etc. on this aspect of Internet history, not all in agreement on the terminology. Perhaps they just used something they thought the audience would relate to. > c. At the time, we would have called it a switch or a packet switch; > internetworking gateway emerged later, and then morphed into a router. An > IMP certainly did routing, but most of us would say that "router" is a > shortened version of "internet router." > d. Ginny, BBN, and others deserve a tremendous amount of credit for > implementing what was then a gateway, now an Internet router (with an > uppercase I) for TCP/IP. As acknowledged by the Computer History Museum: > https://www.computerhistory.org/revolution/networking/19/375/2071 > e. But we often forget that the process was a lot of work, with an > evolution over an extended period of time. > [One small example, just to remind us of what it was like over 40 years > ago, from IEN #25 in 1978: > "The gateways currently use a static routing procedure based on routing > tables assembled into each gateway. In the near future, we plan to > implement a simple gateway routing scheme, which will improve internet > performance by providing the capability to route around failed gateways and > networks." https://www.rfc-editor.org/ien/ien25.pdf To repeat, 1) it was > the best IP/TCP gateway of it's time, 1978, but 2) in order to add a new > network you had to reassemble the code for all gateways, and 3) there was > not yet dynamic routing around a failed network or gateway......] > f. And let's not forget the earlier implementation of an internet gateway > (with a lowercase i) by Ed Taft and Dave Boggs [not me], ca. 1974-1975. > Also from the CHM: > https://www.computerhistory.org/revolution/networking/19/375/2090 The Pup paper you co-authored with Boggs, Taft, and Metcalfe uses the term ?router? in the ?Naming, Addressing, and Routing? section: "This routing process is associated with level 1 in the protocol hierarchy, the level at which packet formats and internet addresses are standardized. The software implementing level 1 is sometimes referred to as a router." ?gregbo From gregskinner0 at icloud.com Wed Nov 24 09:56:44 2021 From: gregskinner0 at icloud.com (Greg Skinner) Date: Wed, 24 Nov 2021 09:56:44 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <1800874110.4972359.1637731162424@mail.yahoo.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> Message-ID: <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> Hi Barbara, I remember those cisco AGS boxes appeared at SRI sometime around the summer of 1986. Some of the SRI staff who maintained them called them ?routers?. As time went on, that name stuck. They may have provided access to BARRnet, the SF bay area NSFnet regional network. ?gregbo > On Nov 23, 2021, at 9:19 PM, Barbara Denny wrote: > > I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. > It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. > barbara > On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote: > > Sorry, I wasn't clear -- the incident I recalled about the sales people > and "gateway" versus "router" was probably a few years later - early > 80s.? We always called them "gateways" until then.? Jack > > On 11/22/21 1:34 PM, vinton cerf wrote: >> your memory and mine are coincident?- i had the impression that >> "router" came from Cisco Systems but that had to be after 1984. If you >> introduced the term "router" in 1977 that would certainly be early. I >> stuck with "gateways" for quite a while after the 1977 three network >> demonstration. >> >> v >> >> >> On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history >> > > wrote: >> >> ? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >> ? ? "clue" which was a picture of Len Kleinrock standing in front of the >> ? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >> ? ? the >> ? ? category "It's a New Machine". >> >> ? ? The host read the clue: >> >> ? ? "In 1969, Leonard Kleinrock and his team sent the first message >> ? ? over the >> ? ? Internet with the first THIS, which now connects devices like modems." >> >> ? ? None of the contestants knew the answer, although one guessed >> ? ? "backbone", which isn't a bad guess.? So the guest revealed the >> ? ? answer: >> >> ? ? "You're looking at the first router." >> >> ? ? My immediate reaction was "No, you're not!??? That's an IMP." >> >> ? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >> ? ? starting at about 1:29 >> >> ? ? That's not quite like I remember it.? Ginny Strazisar built the first >> ? ? switching device for the Internet to connect the ARPANET to the >> ? ? Packet >> ? ? Radio net, circa 1977.? To me that was the genesis of the >> ? ? "Internet" - >> ? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >> ? ? glue >> ? ? it all together.?? But millions of people just learned otherwise. >> >> ? ? A historical tidbit -- Back in 1977 Ginny's system was called a >> ? ? "gateway", but later was renamed a "router".?? It's possible that >> ? ? I did >> ? ? the renaming.? At BBN we were selling lots of packet switches, and >> ? ? sometimes customers asked for ideas on how to use their LANs in the >> ? ? network.? Our sales people would tell them about the research >> ? ? activities, and the role of gateways, TCP/IP , and the Internet. >> ? ? But in >> ? ? many customers' minds that term "gateway"? immediately set off alarm >> ? ? bells, because they had prior bad experience with "gateways" in their >> ? ? IBM networks, and didn't want anything to do with more >> ? ? "gateways".?? So >> ? ? I suggested calling them "routers" instead of "gateways", and >> ? ? suddenly >> ? ? the marketplace became much more willing to listen. >> >> ? ? Enjoy, >> ? ? Jack Haverty >> >> >> ? ? -- >> ? ? Internet-history mailing list >> ? ? Internet-history at elists.isoc.org >> ? ? >> ? ? https://elists.isoc.org/mailman/listinfo/internet-history >> ? ? >> > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > > From sob at sobco.com Wed Nov 24 10:17:19 2021 From: sob at sobco.com (Scott Bradner) Date: Wed, 24 Nov 2021 13:17:19 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> Message-ID: <17B0FBAF-79A5-436F-AD61-B6ECC0DA6D92@sobco.com> fwiw - I have a 1981 Xerox document on "Internet Transport Protocols" - actually about XNS in it Xerox refers to "internetwork routers" so the term was used at least by then Scott > On Nov 24, 2021, at 12:56 PM, Greg Skinner via Internet-history wrote: > > Hi Barbara, > > I remember those cisco AGS boxes appeared at SRI sometime around the summer of 1986. Some of the SRI staff who maintained them called them ?routers?. As time went on, that name stuck. They may have provided access to BARRnet, the SF bay area NSFnet regional network. > > ?gregbo > >> On Nov 23, 2021, at 9:19 PM, Barbara Denny wrote: >> >> I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. >> It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. >> barbara >> On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote: >> >> Sorry, I wasn't clear -- the incident I recalled about the sales people >> and "gateway" versus "router" was probably a few years later - early >> 80s.? We always called them "gateways" until then.? Jack >> >> On 11/22/21 1:34 PM, vinton cerf wrote: >>> your memory and mine are coincident?- i had the impression that >>> "router" came from Cisco Systems but that had to be after 1984. If you >>> introduced the term "router" in 1977 that would certainly be early. I >>> stuck with "gateways" for quite a while after the 1977 three network >>> demonstration. >>> >>> v >>> >>> >>> On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history >>> >> > wrote: >>> >>> ? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >>> ? ? "clue" which was a picture of Len Kleinrock standing in front of the >>> ? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >>> ? ? the >>> ? ? category "It's a New Machine". >>> >>> ? ? The host read the clue: >>> >>> ? ? "In 1969, Leonard Kleinrock and his team sent the first message >>> ? ? over the >>> ? ? Internet with the first THIS, which now connects devices like modems." >>> >>> ? ? None of the contestants knew the answer, although one guessed >>> ? ? "backbone", which isn't a bad guess.? So the guest revealed the >>> ? ? answer: >>> >>> ? ? "You're looking at the first router." >>> >>> ? ? My immediate reaction was "No, you're not!??? That's an IMP." >>> >>> ? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >>> ? ? starting at about 1:29 >>> >>> ? ? That's not quite like I remember it.? Ginny Strazisar built the first >>> ? ? switching device for the Internet to connect the ARPANET to the >>> ? ? Packet >>> ? ? Radio net, circa 1977.? To me that was the genesis of the >>> ? ? "Internet" - >>> ? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >>> ? ? glue >>> ? ? it all together.?? But millions of people just learned otherwise. >>> >>> ? ? A historical tidbit -- Back in 1977 Ginny's system was called a >>> ? ? "gateway", but later was renamed a "router".?? It's possible that >>> ? ? I did >>> ? ? the renaming.? At BBN we were selling lots of packet switches, and >>> ? ? sometimes customers asked for ideas on how to use their LANs in the >>> ? ? network.? Our sales people would tell them about the research >>> ? ? activities, and the role of gateways, TCP/IP , and the Internet. >>> ? ? But in >>> ? ? many customers' minds that term "gateway"? immediately set off alarm >>> ? ? bells, because they had prior bad experience with "gateways" in their >>> ? ? IBM networks, and didn't want anything to do with more >>> ? ? "gateways".?? So >>> ? ? I suggested calling them "routers" instead of "gateways", and >>> ? ? suddenly >>> ? ? the marketplace became much more willing to listen. >>> >>> ? ? Enjoy, >>> ? ? Jack Haverty >>> >>> >>> ? ? -- >>> ? ? Internet-history mailing list >>> ? ? Internet-history at elists.isoc.org >>> ? ? >>> ? ? https://elists.isoc.org/mailman/listinfo/internet-history >>> ? ? >>> >> >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> >> > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From tte at cs.fau.de Wed Nov 24 10:31:37 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Wed, 24 Nov 2021 19:31:37 +0100 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <1800874110.4972359.1637731162424@mail.yahoo.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> Message-ID: Introduction of the term "router" was (IMHO) nevertheless sufficient for a sufficient terminology. If someone wants to use todays strictest definition of "router", then it would include only those features that are architecturally happening up to layer 3 (IP/IPv6). Likewise, a clean "gateway" would only involve any functions that on the gatewya at least include a transport-level function. This leaves out all the rich "hacks" we have introduced over 30 years into routers: NAT ?! (still argued in IETF whether thats a "clean" network layer function, but after 26 IPv4/IPv6 NAT mechanisms...) DPI - Deep Packet inspection, starting from transport port based inspection. TCP fixups such as local rebuffering/retransmissions to overcome window size issues Accounting/Billing based on DPI (e.g.: IPFix, RTP perf-monitoring). Security based on DPI (filtering/ACL, but also more automated data models like MUD or captive portal on router functions). QoS/DiffServ based on DPI "Performance Routing" based on DPI All very useful for business, all very crappy architecturally, no drive to change / clean-up any of this though. *sigh* Cheers Toerless On Wed, Nov 24, 2021 at 05:19:22AM +0000, Barbara Denny via Internet-history wrote: > I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. > It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. > barbara > On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote: > > Sorry, I wasn't clear -- the incident I recalled about the sales people > and "gateway" versus "router" was probably a few years later - early > 80s.? We always called them "gateways" until then.? Jack > > On 11/22/21 1:34 PM, vinton cerf wrote: > > your memory and mine are coincident?- i had the impression that > > "router" came from Cisco Systems but that had to be after 1984. If you > > introduced the term "router" in 1977 that would certainly be early. I > > stuck with "gateways" for quite a while after the 1977 three network > > demonstration. > > > > v > > > > > > On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history > > > > wrote: > > > >? ? Watching Jeopardy from November 17, 2021, I was surprised to see a > >? ? "clue" which was a picture of Len Kleinrock standing in front of the > >? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under > >? ? the > >? ? category "It's a New Machine". > > > >? ? The host read the clue: > > > >? ? "In 1969, Leonard Kleinrock and his team sent the first message > >? ? over the > >? ? Internet with the first THIS, which now connects devices like modems." > > > >? ? None of the contestants knew the answer, although one guessed > >? ? "backbone", which isn't a bad guess.? So the guest revealed the > >? ? answer: > > > >? ? "You're looking at the first router." > > > >? ? My immediate reaction was "No, you're not!??? That's an IMP." > > > >? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw > >? ? starting at about 1:29 > > > >? ? That's not quite like I remember it.? Ginny Strazisar built the first > >? ? switching device for the Internet to connect the ARPANET to the > >? ? Packet > >? ? Radio net, circa 1977.? To me that was the genesis of the > >? ? "Internet" - > >? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to > >? ? glue > >? ? it all together.?? But millions of people just learned otherwise. > > > >? ? A historical tidbit -- Back in 1977 Ginny's system was called a > >? ? "gateway", but later was renamed a "router".?? It's possible that > >? ? I did > >? ? the renaming.? At BBN we were selling lots of packet switches, and > >? ? sometimes customers asked for ideas on how to use their LANs in the > >? ? network.? Our sales people would tell them about the research > >? ? activities, and the role of gateways, TCP/IP , and the Internet. > >? ? But in > >? ? many customers' minds that term "gateway"? immediately set off alarm > >? ? bells, because they had prior bad experience with "gateways" in their > >? ? IBM networks, and didn't want anything to do with more > >? ? "gateways".?? So > >? ? I suggested calling them "routers" instead of "gateways", and > >? ? suddenly > >? ? the marketplace became much more willing to listen. > > > >? ? Enjoy, > >? ? Jack Haverty > > > > > >? ? -- > >? ? Internet-history mailing list > >? ? Internet-history at elists.isoc.org > >? ? > >? ? https://elists.isoc.org/mailman/listinfo/internet-history > >? ? > > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history -- --- tte at cs.fau.de From b_a_denny at yahoo.com Wed Nov 24 11:32:15 2021 From: b_a_denny at yahoo.com (Barbara Denny) Date: Wed, 24 Nov 2021 19:32:15 +0000 (UTC) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> Message-ID: <1059299885.1069268.1637782335529@mail.yahoo.com> The Cisco AGS was released in May 1986 according to Wikipedia. I installed them in Germany for USAEUR probably in September of that year as best as I can recall.? I think the deployment first included IMPs but I think subsequently SRI removed the IMPs that were between the Cisco boxes (though I think IMPs we're called PSNs by this point in time. BBN certainly can chime in here).? I don't know about? SRI people using Cisco's AGS boxes for something like the bay area regional network.? barbara On Wednesday, November 24, 2021, 09:56:48 AM PST, Greg Skinner wrote: Hi Barbara, I remember those cisco AGS boxes appeared at SRI sometime around the summer of 1986.? Some of the SRI staff who maintained them called them ?routers?.? As time went on, that name stuck.? They may have provided access to BARRnet, the SF bay area NSFnet regional network. ?gregbo > On Nov 23, 2021, at 9:19 PM, Barbara Denny wrote: > > I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. > It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. > barbara >? ? On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote:? > > Sorry, I wasn't clear -- the incident I recalled about the sales people > and "gateway" versus "router" was probably a few years later - early > 80s.? We always called them "gateways" until then.? Jack > > On 11/22/21 1:34 PM, vinton cerf wrote: >> your memory and mine are coincident?- i had the impression that >> "router" came from Cisco Systems but that had to be after 1984. If you >> introduced the term "router" in 1977 that would certainly be early. I >> stuck with "gateways" for quite a while after the 1977 three network >> demonstration. >> >> v >> >> >> On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history >> > > wrote: >> >> ? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >> ? ? "clue" which was a picture of Len Kleinrock standing in front of the >> ? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >> ? ? the >> ? ? category "It's a New Machine". >> >> ? ? The host read the clue: >> >> ? ? "In 1969, Leonard Kleinrock and his team sent the first message >> ? ? over the >> ? ? Internet with the first THIS, which now connects devices like modems." >> >> ? ? None of the contestants knew the answer, although one guessed >> ? ? "backbone", which isn't a bad guess.? So the guest revealed the >> ? ? answer: >> >> ? ? "You're looking at the first router." >> >> ? ? My immediate reaction was "No, you're not!??? That's an IMP." >> >> ? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >> ? ? starting at about 1:29 >> >> ? ? That's not quite like I remember it.? Ginny Strazisar built the first >> ? ? switching device for the Internet to connect the ARPANET to the >> ? ? Packet >> ? ? Radio net, circa 1977.? To me that was the genesis of the >> ? ? "Internet" - >> ? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >> ? ? glue >> ? ? it all together.?? But millions of people just learned otherwise. >> >> ? ? A historical tidbit -- Back in 1977 Ginny's system was called a >> ? ? "gateway", but later was renamed a "router".?? It's possible that >> ? ? I did >> ? ? the renaming.? At BBN we were selling lots of packet switches, and >> ? ? sometimes customers asked for ideas on how to use their LANs in the >> ? ? network.? Our sales people would tell them about the research >> ? ? activities, and the role of gateways, TCP/IP , and the Internet. >> ? ? But in >> ? ? many customers' minds that term "gateway"? immediately set off alarm >> ? ? bells, because they had prior bad experience with "gateways" in their >> ? ? IBM networks, and didn't want anything to do with more >> ? ? "gateways".?? So >> ? ? I suggested calling them "routers" instead of "gateways", and >> ? ? suddenly >> ? ? the marketplace became much more willing to listen. >> >> ? ? Enjoy, >> ? ? Jack Haverty >> >> >> ? ? -- >> ? ? Internet-history mailing list >> ? ? Internet-history at elists.isoc.org >> ? ? >> ? ? https://elists.isoc.org/mailman/listinfo/internet-history >> ? ? >> > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > > From b_a_denny at yahoo.com Wed Nov 24 11:36:35 2021 From: b_a_denny at yahoo.com (Barbara Denny) Date: Wed, 24 Nov 2021 19:36:35 +0000 (UTC) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <1059299885.1069268.1637782335529@mail.yahoo.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> <1059299885.1069268.1637782335529@mail.yahoo.com> Message-ID: <968246566.1070688.1637782595509@mail.yahoo.com> Sorry folks. That should have been USAREUR.? I have trouble keeping up with auto correction on my phone. Happy Thanksgiving! barbara On Wednesday, November 24, 2021, 11:32:41 AM PST, Barbara Denny via Internet-history wrote: The Cisco AGS was released in May 1986 according to Wikipedia. I installed them in Germany for USAEUR probably in September of that year as best as I can recall.? I think the deployment first included IMPs but I think subsequently SRI removed the IMPs that were between the Cisco boxes (though I think IMPs we're called PSNs by this point in time. BBN certainly can chime in here).? I don't know about? SRI people using Cisco's AGS boxes for something like the bay area regional network.? barbara ? ? On Wednesday, November 24, 2021, 09:56:48 AM PST, Greg Skinner wrote:? Hi Barbara, I remember those cisco AGS boxes appeared at SRI sometime around the summer of 1986.? Some of the SRI staff who maintained them called them ?routers?.? As time went on, that name stuck.? They may have provided access to BARRnet, the SF bay area NSFnet regional network. ?gregbo > On Nov 23, 2021, at 9:19 PM, Barbara Denny wrote: > > I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. > It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. > barbara >? ? On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote:? > > Sorry, I wasn't clear -- the incident I recalled about the sales people > and "gateway" versus "router" was probably a few years later - early > 80s.? We always called them "gateways" until then.? Jack > > On 11/22/21 1:34 PM, vinton cerf wrote: >> your memory and mine are coincident?- i had the impression that >> "router" came from Cisco Systems but that had to be after 1984. If you >> introduced the term "router" in 1977 that would certainly be early. I >> stuck with "gateways" for quite a while after the 1977 three network >> demonstration. >> >> v >> >> >> On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history >> > > wrote: >> >> ? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >> ? ? "clue" which was a picture of Len Kleinrock standing in front of the >> ? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >> ? ? the >> ? ? category "It's a New Machine". >> >> ? ? The host read the clue: >> >> ? ? "In 1969, Leonard Kleinrock and his team sent the first message >> ? ? over the >> ? ? Internet with the first THIS, which now connects devices like modems." >> >> ? ? None of the contestants knew the answer, although one guessed >> ? ? "backbone", which isn't a bad guess.? So the guest revealed the >> ? ? answer: >> >> ? ? "You're looking at the first router." >> >> ? ? My immediate reaction was "No, you're not!??? That's an IMP." >> >> ? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >> ? ? starting at about 1:29 >> >> ? ? That's not quite like I remember it.? Ginny Strazisar built the first >> ? ? switching device for the Internet to connect the ARPANET to the >> ? ? Packet >> ? ? Radio net, circa 1977.? To me that was the genesis of the >> ? ? "Internet" - >> ? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >> ? ? glue >> ? ? it all together.?? But millions of people just learned otherwise. >> >> ? ? A historical tidbit -- Back in 1977 Ginny's system was called a >> ? ? "gateway", but later was renamed a "router".?? It's possible that >> ? ? I did >> ? ? the renaming.? At BBN we were selling lots of packet switches, and >> ? ? sometimes customers asked for ideas on how to use their LANs in the >> ? ? network.? Our sales people would tell them about the research >> ? ? activities, and the role of gateways, TCP/IP , and the Internet. >> ? ? But in >> ? ? many customers' minds that term "gateway"? immediately set off alarm >> ? ? bells, because they had prior bad experience with "gateways" in their >> ? ? IBM networks, and didn't want anything to do with more >> ? ? "gateways".?? So >> ? ? I suggested calling them "routers" instead of "gateways", and >> ? ? suddenly >> ? ? the marketplace became much more willing to listen. >> >> ? ? Enjoy, >> ? ? Jack Haverty >> >> >> ? ? -- >> ? ? Internet-history mailing list >> ? ? Internet-history at elists.isoc.org >> ? ? >> ? ? https://elists.isoc.org/mailman/listinfo/internet-history >> ? ? >> > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > > ? -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From jack at 3kitty.org Wed Nov 24 13:02:06 2021 From: jack at 3kitty.org (Jack Haverty) Date: Wed, 24 Nov 2021 13:02:06 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <968246566.1070688.1637782595509@mail.yahoo.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> <1059299885.1069268.1637782335529@mail.yahoo.com> <968246566.1070688.1637782595509@mail.yahoo.com> Message-ID: IMHO, it's usually not possible to agree on the meaning of network terminology, or readily reconstruct the history of its changes. There is no universal "standards body" dictating what terms mean. Terms mean what people think they mean, at the time, and they don't typically agree.? I learned in high school debating that the first step in a debate is to "define your terms" so that the ensuing debate can at least have everyone, in that debate, talking about the same things, at least for the next hour or so. Today, there is no universal agreement on networking terms -- even including words such as "Internet".?? It's not a new situation. Back in the early 70s, the ARPANET IMPs sent and received "messages" to and from host computers, and split them up into "packets" for transmission between IMPs.? But as host people, we thought we were sending "packets" to the IMP, especially after "message" got to mean mail objects to many people.? When TCP appeared, so did datagrams - but IIRC even that term was already is use somewhere else. So, my anecdote about changing from "gateway" to "router" at BBN in the early 80s was when I redefined my personal terminology (and the BBN sales people adopted the new term eagerly so as not to lose customers' interest).? It's quite possible I heard the term "router" earlier in the context of PARC's work (but I declined to call packets "pups", another term that has died out).? One of the Internet meetings in the early 80s was held at PARC and I'm sure we got a presentation on how Xerox was building their own Internet. What did they call it other than XNS?? Was it "the Xerox internet"? How about "internets" built using Netware, OSI, LU6.2, DECNET, Appletalk, etc.?? What terms did they use?? I can't recall. Even email arguably also involved routing, and used terms such as "gateway".?? It even had its own interconnection of networks, more extensive even than "the Internet", and used a form of Source Routing to allow senders to specify how a message should be routed (remember ! % and such in mail addresses?).? AFAIK mail servers were never called routers -- but that's part of what they did and still do.?? You can see this in the collection of fields in headers. Terms change over time as people make decisions.? I think these are often "marketing" decisions, even when made by techies.? If you choose to present your idea as "New!? Improved!", you call it the same as something older, and explain why yours is better.?? TCP/IP V4 is of course better than V2, and V6 is even better.? If you choose to present your idea as "New! Finally Solves Your Problems!", you give it a new name and explain why that old stuff is plainly wrong and should be discarded immediately.?? TCP/IP is of course better than X.25/X.75. When I was talking with a lawyer once during a patent battle, he observed that most of a court's work involves arguing about the meaning of words, which can change significantly over the years. /Jack On 11/24/21 11:36 AM, Barbara Denny via Internet-history wrote: > Sorry folks. That should have been USAREUR.? I have trouble keeping up with auto correction on my phone. > Happy Thanksgiving! > barbara > On Wednesday, November 24, 2021, 11:32:41 AM PST, Barbara Denny via Internet-history wrote: > > > The Cisco AGS was released in May 1986 according to Wikipedia. I installed them in Germany for USAEUR probably in September of that year as best as I can recall.? I think the deployment first included IMPs but I think subsequently SRI removed the IMPs that were between the Cisco boxes (though I think IMPs we're called PSNs by this point in time. BBN certainly can chime in here).? I don't know about? SRI people using Cisco's AGS boxes for something like the bay area regional network. > barbara > > ? ? On Wednesday, November 24, 2021, 09:56:48 AM PST, Greg Skinner wrote: > > Hi Barbara, > > I remember those cisco AGS boxes appeared at SRI sometime around the summer of 1986.? Some of the SRI staff who maintained them called them ?routers?.? As time went on, that name stuck.? They may have provided access to BARRnet, the SF bay area NSFnet regional network. > > ?gregbo > >> On Nov 23, 2021, at 9:19 PM, Barbara Denny wrote: >> >> I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. >> It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. >> barbara >> ? ? On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote: >> >> Sorry, I wasn't clear -- the incident I recalled about the sales people >> and "gateway" versus "router" was probably a few years later - early >> 80s.? We always called them "gateways" until then.? Jack >> >> On 11/22/21 1:34 PM, vinton cerf wrote: >>> your memory and mine are coincident?- i had the impression that >>> "router" came from Cisco Systems but that had to be after 1984. If you >>> introduced the term "router" in 1977 that would certainly be early. I >>> stuck with "gateways" for quite a while after the 1977 three network >>> demonstration. >>> >>> v >>> >>> >>> On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history >>> >> > wrote: >>> >>> ? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >>> ? ? "clue" which was a picture of Len Kleinrock standing in front of the >>> ? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >>> ? ? the >>> ? ? category "It's a New Machine". >>> >>> ? ? The host read the clue: >>> >>> ? ? "In 1969, Leonard Kleinrock and his team sent the first message >>> ? ? over the >>> ? ? Internet with the first THIS, which now connects devices like modems." >>> >>> ? ? None of the contestants knew the answer, although one guessed >>> ? ? "backbone", which isn't a bad guess.? So the guest revealed the >>> ? ? answer: >>> >>> ? ? "You're looking at the first router." >>> >>> ? ? My immediate reaction was "No, you're not!??? That's an IMP." >>> >>> ? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >>> ? ? starting at about 1:29 >>> >>> ? ? That's not quite like I remember it.? Ginny Strazisar built the first >>> ? ? switching device for the Internet to connect the ARPANET to the >>> ? ? Packet >>> ? ? Radio net, circa 1977.? To me that was the genesis of the >>> ? ? "Internet" - >>> ? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >>> ? ? glue >>> ? ? it all together.?? But millions of people just learned otherwise. >>> >>> ? ? A historical tidbit -- Back in 1977 Ginny's system was called a >>> ? ? "gateway", but later was renamed a "router".?? It's possible that >>> ? ? I did >>> ? ? the renaming.? At BBN we were selling lots of packet switches, and >>> ? ? sometimes customers asked for ideas on how to use their LANs in the >>> ? ? network.? Our sales people would tell them about the research >>> ? ? activities, and the role of gateways, TCP/IP , and the Internet. >>> ? ? But in >>> ? ? many customers' minds that term "gateway"? immediately set off alarm >>> ? ? bells, because they had prior bad experience with "gateways" in their >>> ? ? IBM networks, and didn't want anything to do with more >>> ? ? "gateways".?? So >>> ? ? I suggested calling them "routers" instead of "gateways", and >>> ? ? suddenly >>> ? ? the marketplace became much more willing to listen. >>> >>> ? ? Enjoy, >>> ? ? Jack Haverty >>> >>> >>> ? ? -- >>> ? ? Internet-history mailing list >>> ? ? Internet-history at elists.isoc.org >>> ? ? >>> ? ? https://elists.isoc.org/mailman/listinfo/internet-history >>> ? ? >>> >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> >> > From brian.e.carpenter at gmail.com Wed Nov 24 13:14:15 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Thu, 25 Nov 2021 10:14:15 +1300 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <17B0FBAF-79A5-436F-AD61-B6ECC0DA6D92@sobco.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> <17B0FBAF-79A5-436F-AD61-B6ECC0DA6D92@sobco.com> Message-ID: <83074788-d46c-6202-0989-1ab259c8b40b@gmail.com> That made me think of one of the famous PARC blue-and-whites. I no longer have my copy, but Google found it. "The software implementing level 1 is sometimes referred to as a router." Page 7 of "Pup: An Internetwork Architecture" by Boggs, Shoch, Taft, Metcalfe, Xerox PARC CSL-79-10, July 1979. (http://www.bitsavers.org/pdf/xerox/parc/techReports/CSL-79-10_Pup_An_Internetwork_Architecture_Jul79.pdf) Naturally, and correctly, the document uses "internet" (lower case) as a generic term. Note that IEN19 (dated January 1978) does not use "router". It comes very close, but sticks to "gateway" as the noun. Since John Shoch is on this list, maybe he remembers where the term came from, but it looks likely that it originated at 3333 Coyote Hill Road in 1978-79. I also checked "Computer Networks and their Protocols" (Davies et al, 1979) and that also sticks to "gateway". Regards Brian On 25-Nov-21 07:17, Scott Bradner via Internet-history wrote: > fwiw - I have a 1981 Xerox document on "Internet Transport Protocols" - actually about XNS > > in it Xerox refers to "internetwork routers" so the term was used at least by then > > Scott > >> On Nov 24, 2021, at 12:56 PM, Greg Skinner via Internet-history wrote: >> >> Hi Barbara, >> >> I remember those cisco AGS boxes appeared at SRI sometime around the summer of 1986. Some of the SRI staff who maintained them called them ?routers?. As time went on, that name stuck. They may have provided access to BARRnet, the SF bay area NSFnet regional network. >> >> ?gregbo >> >>> On Nov 23, 2021, at 9:19 PM, Barbara Denny wrote: >>> >>> I think the change from router to gateway in the networking community was suggested somewhere in the 1983-1986 timeframe .? I am using the final report for the Reconstitution Protocol project? to jog my memory.? ?I remember Jim Mathis telling me about this change in terminology because there was at least one other device, called a gateway, that provided a translation like functionality and people didn't want to create more confusion.? I also don't think we would have used the word gateway during the RP project if router was the accepted term when we started. I am pretty sure we also were using the AGS Cisco boxes? in? the fall of 1986 for a different SRI project so I think the change in naming was?underway by then. I don't remember feeling any surprise at Cisco's use of the word router to describe their product. >>> It is interesting to remember that BGP as an experimental RFC wasn't released until 1989 so we still had the term gateway hanging around for quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated April 1987. >>> barbara >>> On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via Internet-history wrote: >>> >>> Sorry, I wasn't clear -- the incident I recalled about the sales people >>> and "gateway" versus "router" was probably a few years later - early >>> 80s.? We always called them "gateways" until then.? Jack >>> >>> On 11/22/21 1:34 PM, vinton cerf wrote: >>>> your memory and mine are coincident?- i had the impression that >>>> "router" came from Cisco Systems but that had to be after 1984. If you >>>> introduced the term "router" in 1977 that would certainly be early. I >>>> stuck with "gateways" for quite a while after the 1977 three network >>>> demonstration. >>>> >>>> v >>>> >>>> >>>> On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history >>>> >>> > wrote: >>>> >>>> ? ? Watching Jeopardy from November 17, 2021, I was surprised to see a >>>> ? ? "clue" which was a picture of Len Kleinrock standing in front of the >>>> ? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under >>>> ? ? the >>>> ? ? category "It's a New Machine". >>>> >>>> ? ? The host read the clue: >>>> >>>> ? ? "In 1969, Leonard Kleinrock and his team sent the first message >>>> ? ? over the >>>> ? ? Internet with the first THIS, which now connects devices like modems." >>>> >>>> ? ? None of the contestants knew the answer, although one guessed >>>> ? ? "backbone", which isn't a bad guess.? So the guest revealed the >>>> ? ? answer: >>>> >>>> ? ? "You're looking at the first router." >>>> >>>> ? ? My immediate reaction was "No, you're not!??? That's an IMP." >>>> >>>> ? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw >>>> ? ? starting at about 1:29 >>>> >>>> ? ? That's not quite like I remember it.? Ginny Strazisar built the first >>>> ? ? switching device for the Internet to connect the ARPANET to the >>>> ? ? Packet >>>> ? ? Radio net, circa 1977.? To me that was the genesis of the >>>> ? ? "Internet" - >>>> ? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to >>>> ? ? glue >>>> ? ? it all together.?? But millions of people just learned otherwise. >>>> >>>> ? ? A historical tidbit -- Back in 1977 Ginny's system was called a >>>> ? ? "gateway", but later was renamed a "router".?? It's possible that >>>> ? ? I did >>>> ? ? the renaming.? At BBN we were selling lots of packet switches, and >>>> ? ? sometimes customers asked for ideas on how to use their LANs in the >>>> ? ? network.? Our sales people would tell them about the research >>>> ? ? activities, and the role of gateways, TCP/IP , and the Internet. >>>> ? ? But in >>>> ? ? many customers' minds that term "gateway"? immediately set off alarm >>>> ? ? bells, because they had prior bad experience with "gateways" in their >>>> ? ? IBM networks, and didn't want anything to do with more >>>> ? ? "gateways".?? So >>>> ? ? I suggested calling them "routers" instead of "gateways", and >>>> ? ? suddenly >>>> ? ? the marketplace became much more willing to listen. >>>> >>>> ? ? Enjoy, >>>> ? ? Jack Haverty >>>> >>>> >>>> ? ? -- >>>> ? ? Internet-history mailing list >>>> ? ? Internet-history at elists.isoc.org >>>> ? ? >>>> ? ? https://elists.isoc.org/mailman/listinfo/internet-history >>>> ? ? >>>> >>> >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history >>> >>> >> >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history > From vint at google.com Wed Nov 24 13:16:28 2021 From: vint at google.com (Vint Cerf) Date: Wed, 24 Nov 2021 16:16:28 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <83074788-d46c-6202-0989-1ab259c8b40b@gmail.com> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <589872a7-0f29-f449-637c-1cb5be7ff5a7@3kitty.org> <1800874110.4972359.1637731162424@mail.yahoo.com> <67C57EF7-935E-46ED-93BF-6273A4500C9F@icloud.com> <17B0FBAF-79A5-436F-AD61-B6ECC0DA6D92@sobco.com> <83074788-d46c-6202-0989-1ab259c8b40b@gmail.com> Message-ID: as email was gearing up but was not standardized, we used to refer to "application layer gateways" to explain how email connectivity worked between the Internet and other systems. v On Wed, Nov 24, 2021 at 4:14 PM Brian E Carpenter via Internet-history < internet-history at elists.isoc.org> wrote: > That made me think of one of the famous PARC blue-and-whites. I no longer > have my copy, but Google found it. > > "The software implementing level 1 is sometimes referred to as a router." > > Page 7 of "Pup: An Internetwork Architecture" by Boggs, Shoch, Taft, > Metcalfe, Xerox PARC CSL-79-10, July 1979. ( > http://www.bitsavers.org/pdf/xerox/parc/techReports/CSL-79-10_Pup_An_Internetwork_Architecture_Jul79.pdf > ) > > Naturally, and correctly, the document uses "internet" (lower case) as a > generic term. > > Note that IEN19 (dated January 1978) does not use "router". It comes very > close, but sticks to "gateway" as the noun. Since John Shoch is on this > list, maybe he remembers where the term came from, but it looks likely that > it originated at 3333 Coyote Hill Road in 1978-79. > > I also checked "Computer Networks and their Protocols" (Davies et al, > 1979) and that also sticks to "gateway". > > Regards > Brian > > On 25-Nov-21 07:17, Scott Bradner via Internet-history wrote: > > fwiw - I have a 1981 Xerox document on "Internet Transport Protocols" - > actually about XNS > > > > in it Xerox refers to "internetwork routers" so the term was used at > least by then > > > > Scott > > > >> On Nov 24, 2021, at 12:56 PM, Greg Skinner via Internet-history < > internet-history at elists.isoc.org> wrote: > >> > >> Hi Barbara, > >> > >> I remember those cisco AGS boxes appeared at SRI sometime around the > summer of 1986. Some of the SRI staff who maintained them called them > ?routers?. As time went on, that name stuck. They may have > provided access to BARRnet, the SF bay area NSFnet regional network. > >> > >> ?gregbo > >> > >>> On Nov 23, 2021, at 9:19 PM, Barbara Denny > wrote: > >>> > >>> I think the change from router to gateway in the networking community > was suggested somewhere in the 1983-1986 timeframe .? I am using the final > report for the Reconstitution Protocol project? to jog my memory.? ?I > remember Jim Mathis telling me about this change in terminology because > there was at least one other device, called a gateway, that provided a > translation like functionality and people didn't want to create more > confusion.? I also don't think we would have used the word gateway during > the RP project if router was the accepted term when we started. I am pretty > sure > we also were using the AGS Cisco boxes? in? the fall of 1986 for a > different SRI project so I think the change in naming was?underway by then. > I don't remember feeling any surprise at Cisco's use of the word router to > describe their product. > >>> It is interesting to remember that BGP as an experimental RFC wasn't > released until 1989 so we still had the term gateway hanging around for > quite a while afterwards.? I don't mean to ignore IGRP and EIGRP either. > Then there is also Lixia Zhang's? paper on "How to Build a Gateway" dated > April 1987. > >>> barbara > >>> On Tuesday, November 23, 2021, 10:39:31 AM PST, Jack Haverty via > Internet-history wrote: > >>> > >>> Sorry, I wasn't clear -- the incident I recalled about the sales people > >>> and "gateway" versus "router" was probably a few years later - early > >>> 80s.? We always called them "gateways" until then.? Jack > >>> > >>> On 11/22/21 1:34 PM, vinton cerf wrote: > >>>> your memory and mine are coincident?- i had the impression that > >>>> "router" came from Cisco Systems but that had to be after 1984. If you > >>>> introduced the term "router" in 1977 that would certainly be early. I > >>>> stuck with "gateways" for quite a while after the 1977 three network > >>>> demonstration. > >>>> > >>>> v > >>>> > >>>> > >>>> On Mon, Nov 22, 2021 at 3:50 PM Jack Haverty via Internet-history > >>>> >>>> > wrote: > >>>> > >>>> ? ? Watching Jeopardy from November 17, 2021, I was surprised to see > a > >>>> ? ? "clue" which was a picture of Len Kleinrock standing in front of > the > >>>> ? ? ARPANET IMP which has been preserved at UCLA.? It was a clue under > >>>> ? ? the > >>>> ? ? category "It's a New Machine". > >>>> > >>>> ? ? The host read the clue: > >>>> > >>>> ? ? "In 1969, Leonard Kleinrock and his team sent the first message > >>>> ? ? over the > >>>> ? ? Internet with the first THIS, which now connects devices like > modems." > >>>> > >>>> ? ? None of the contestants knew the answer, although one guessed > >>>> ? ? "backbone", which isn't a bad guess.? So the guest revealed the > >>>> ? ? answer: > >>>> > >>>> ? ? "You're looking at the first router." > >>>> > >>>> ? ? My immediate reaction was "No, you're not!??? That's an IMP." > >>>> > >>>> ? ? See https://www.youtube.com/watch?v=Lmv1WKMGcDw > >>>> ? ? starting at about > 1:29 > >>>> > >>>> ? ? That's not quite like I remember it.? Ginny Strazisar built the > first > >>>> ? ? switching device for the Internet to connect the ARPANET to the > >>>> ? ? Packet > >>>> ? ? Radio net, circa 1977.? To me that was the genesis of the > >>>> ? ? "Internet" - > >>>> ? ? INTERconnecting of NETworks of a variety of types, using TCP/IP to > >>>> ? ? glue > >>>> ? ? it all together.?? But millions of people just learned otherwise. > >>>> > >>>> ? ? A historical tidbit -- Back in 1977 Ginny's system was called a > >>>> ? ? "gateway", but later was renamed a "router".?? It's possible that > >>>> ? ? I did > >>>> ? ? the renaming.? At BBN we were selling lots of packet switches, and > >>>> ? ? sometimes customers asked for ideas on how to use their LANs in > the > >>>> ? ? network.? Our sales people would tell them about the research > >>>> ? ? activities, and the role of gateways, TCP/IP , and the Internet. > >>>> ? ? But in > >>>> ? ? many customers' minds that term "gateway"? immediately set off > alarm > >>>> ? ? bells, because they had prior bad experience with "gateways" in > their > >>>> ? ? IBM networks, and didn't want anything to do with more > >>>> ? ? "gateways".?? So > >>>> ? ? I suggested calling them "routers" instead of "gateways", and > >>>> ? ? suddenly > >>>> ? ? the marketplace became much more willing to listen. > >>>> > >>>> ? ? Enjoy, > >>>> ? ? Jack Haverty > >>>> > >>>> > >>>> ? ? -- > >>>> ? ? Internet-history mailing list > >>>> ? ? Internet-history at elists.isoc.org > >>>> ? ? > >>>> ? ? https://elists.isoc.org/mailman/listinfo/internet-history > >>>> ? ? > >>>> > >>> > >>> -- > >>> Internet-history mailing list > >>> Internet-history at elists.isoc.org > >>> https://elists.isoc.org/mailman/listinfo/internet-history > >>> > >>> > >> > >> -- > >> Internet-history mailing list > >> Internet-history at elists.isoc.org > >> https://elists.isoc.org/mailman/listinfo/internet-history > > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > -- Please send any postal/overnight deliveries to: Vint Cerf 1435 Woodhurst Blvd McLean, VA 22102 703-448-0965 until further notice From jnc at mercury.lcs.mit.edu Wed Nov 24 14:06:35 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Wed, 24 Nov 2021 17:06:35 -0500 (EST) Subject: [ih] multi-protocol routers, bridges (Was: "The First Router" on Jeopardy) Message-ID: <20211124220636.03A4B18C077@mercury.lcs.mit.edu> > From: Clem Cole > IP had not yet 'won' and a lot of site had multiple protocols running > on their LANs. MIT had that problem in spades, which was exactly the genesis of the multi-protocol router at MIT. CHAOS had a large lead in deployment at MIT, since it got rolling before TCP/IP really did; I think the AI Lab guys got word of what was happening at PARC, and decided to build a rough copy (hardware and, less identically, protcol - the latter eventually ran over DIX Ethernet too). Once LCS got rolling with the rings and TCP/IP, it clearly made sense to have all the LANs carry both (or we'd have to rub both kinds of wire everwhere; much more cost effective to carry both). After Dave Clark and I chatted with Dave Moon for a while, we came with MUPPETs (MIT Universal Packets), to allow both CHAOS and TCP/IP packets on the MIT LANs to share a common carriage format (and thus only one kind of router). The CHAOS guys didn't show much movement towards actually implementhing them, though (no surprise, not much incentive). So then it looked like carrying each protocol independently would be easier to accomplish. That still left the problem of inter-LAN gateways (routers), though - did those have to be replicated? I came up with the multi-protocol router as a way to economize on them. Dave Clark wasn't totally sold on the idea to begin with (he was in charge of prodcing an MIT-wide campus LAN plan at the time, and included the 'multi-protocol spine' as a possible solution, but without saying 'the only plausible solution'). So I set off to write one, to show that it was viable. Somewhere in there the Xerox grant showed up, so we had Experimental Ethernet and PUP to deal with too. (My first kludgy router, written in MACRO-11, handled IP packets over the rings, the Experimental Ethernet and CHAOSNET too, but IP-only, IIRC - I ran across the code the other day, I could look.) (The AI guys actually implemented a service gateway to get to the Dover printer; the AI-CHAOS-11 spoke CHAOS protocol to CHAOSNET hosts, and EFTP to the Dover spooler. They had CHAOS protocol to NCP - later TCP - gateways too.) Something similar for Bill Yeager at Stanford, who independently re-invented the multi-protocol spine/router idea; PUP and IP for them. > Some were routable, some were not. Interesting point, but not true of MIT (and probably Stanford too); all our protocols were router-able. That may be part of the latter attraction of bridges, though. Note that bridges only _really_ work well on LANs with large, unique interface network addresses; only after DIX Ethernet was that true. Early LANs (Proteon rings, Omninet, etc, etc) had small addresses, and had to use routers. > But mixed protocols was probably more the norm in commercials and I bet > University circles than not. I'm not sure. It probably depended on what kind of boxes any particular place had; not sure there was a commercial/academic difference, but I agree on the 'multi-protocol was more common than single rotocol'. > I know an early 3Com brouter but my memory is there were others. I > think DEC made one or two. Noel can tell us what Proeton did. I know that at the start Proteon totally missed the bridge boat. I have to hold up my hand for a lot of that; I was so focused on my 'a bridge looks like a wire, but isn't' mantra (e.g. I thought we'd have to deal with congestion explicitly - this was before Van Jacobsen), I missed out on the advantages of bridges: trivial to install (especially with 'unique at manufacture' interface network addresses), they supported non-routable protocols, etc. Also, with Proteon's own LANs not having unique interface network addresses, we weren't well-positioned to see the ease of use, etc, points. I think they did eventually add bridging functionality towards the end, but I was not very involved there by then. Noel From winowicki at yahoo.com Wed Nov 24 15:06:22 2021 From: winowicki at yahoo.com (Bill Nowicki) Date: Wed, 24 Nov 2021 23:06:22 +0000 (UTC) Subject: [ih] multi-protocol routers, bridges (Was: "The First Router" on Jeopardy) In-Reply-To: <20211124220636.03A4B18C077@mercury.lcs.mit.edu> References: <20211124220636.03A4B18C077@mercury.lcs.mit.edu> Message-ID: <199954801.2108729.1637795183017@mail.yahoo.com> Thanks Noel. Here are a few anecdotes from the opposite coast.For amusement if nothing else. At Stanford we first had an application-level gateway (telnet and mail) so we called the first routers "gateways" too. It was just a PUP and IP gateway. Especially because at first the topology was simple, and routing was easier: what comes in one side goes out the other for example (or we had a tree topology). As we soon got the full cross-product and cloned the MIT LSI-11 system, the term "router" was more often used. One reluctance came from how to pronounce the word. Most of us rhymed it with "outer". However, Steve Deering spoke Canadian, so pronounced it like "rooter". The IMPs of course ran sophisticated routing from the start, so calling them a router seems not so bad, compared to a "broadband solution" or some other meaningless marketing term. The other topic in the original thread was type-of-service routing. We worked on this later when I was at SGI in the 1990s.? We were doing real-time audio and video over a shared corporate IP network pretty heavily. However, the Netscape (remember them?) hype got the attention of executives when they stock took off. SGI disbanded all its Internet investment, and when we asked why, the response was something like: "we will put the video on the web, since Netscape already does that".? A few years later in 2001 I was at a start-up trying to storage over IP using the iSCSI protocol ahead of our time. Entrenched vendors (Sun and Cisco etc. by then) told their customers one could never use IP or Ethernet for storage because they were "unreliable" compared to Fibre Channel. The other two iSCSI vendors were acquired but we went out of business. Then Cisco came out with an Ethernet switch that had two priorities, and called it "lossless Ethernet". It was then acceptable for enterprise customers to start using IP over Ethernet for storage, and of course now days it is used heavily in all the commercial clouds. Bill Nowicki On Wednesday, November 24, 2021, 02:06:43 PM PST, Noel Chiappa via Internet-history wrote: ? ? > From: Clem Cole ? ? > IP had not yet 'won' and a lot of site had multiple protocols running ? ? > on their LANs. MIT had that problem in spades, which was exactly the genesis of the multi-protocol router at MIT. CHAOS had a large lead in deployment at MIT, since it got rolling before TCP/IP really did; I think the AI Lab guys got word of what was happening at PARC, and decided to build a rough copy (hardware and, less identically, protcol - the latter eventually ran over DIX Ethernet too). Once LCS got rolling with the rings and TCP/IP, it clearly made sense to have all the LANs carry both (or we'd have to rub both kinds of wire everwhere; much more cost effective to carry both). After Dave Clark and I chatted with Dave Moon for a while, we came with MUPPETs (MIT Universal Packets), to allow both CHAOS and TCP/IP packets on the MIT LANs to share a common carriage format (and thus only one kind of router). The CHAOS guys didn't show much movement towards actually implementhing them, though (no surprise, not much incentive). So then it looked like carrying each protocol independently would be easier to accomplish. That still left the problem of inter-LAN gateways (routers), though - did those have to be replicated? I came up with the multi-protocol router as a way to economize on them. Dave Clark wasn't totally sold on the idea to begin with (he was in charge of prodcing an MIT-wide campus LAN plan at the time, and included the 'multi-protocol spine' as a possible solution, but without saying 'the only plausible solution'). So I set off to write one, to show that it was viable. Somewhere in there the Xerox grant showed up, so we had Experimental Ethernet and PUP to deal with too. (My first kludgy router, written in MACRO-11, handled IP packets over the rings, the Experimental Ethernet and CHAOSNET too, but IP-only, IIRC - I ran across the code the other day, I could look.) (The AI guys actually implemented a service gateway to get to the Dover printer; the AI-CHAOS-11 spoke CHAOS protocol to CHAOSNET hosts, and EFTP to the Dover spooler. They had CHAOS protocol to NCP - later TCP - gateways too.) Something similar for Bill Yeager at Stanford, who independently re-invented the multi-protocol spine/router idea; PUP and IP for them. ? > Some were routable, some were not. Interesting point, but not true of MIT (and probably Stanford too); all our protocols were router-able. That may be part of the latter attraction of bridges, though. Note that bridges only _really_ work well on LANs with large, unique interface network addresses; only after DIX Ethernet was that true. Early LANs (Proteon rings, Omninet, etc, etc) had small addresses, and had to use routers. ? ? > But mixed protocols was probably more the norm in commercials and I bet ? ? > University circles than not. I'm not sure. It probably depended on what kind of boxes any particular place had; not sure there was a commercial/academic difference, but I agree on the 'multi-protocol was more common than single rotocol'. ? ? > I know an early 3Com brouter but my memory is there were others. I ? ? > think DEC made one or two. Noel can tell us what Proeton did. I know that at the start Proteon totally missed the bridge boat. I have to hold up my hand for a lot of that; I was so focused on my 'a bridge looks like a wire, but isn't' mantra (e.g. I thought we'd have to deal with congestion explicitly - this was before Van Jacobsen), I missed out on the advantages of bridges: trivial to install (especially with 'unique at manufacture' interface network addresses), they supported non-routable protocols, etc. Also, with Proteon's own LANs not having unique interface network addresses, we weren't well-positioned to see the ease of use, etc, points. I think they did eventually add bridging functionality towards the end, but I was not very involved there by then. ??? Noel -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From agmalis at gmail.com Wed Nov 24 15:25:01 2021 From: agmalis at gmail.com (Andrew G. Malis) Date: Wed, 24 Nov 2021 18:25:01 -0500 Subject: [ih] For our next argument .... Message-ID: Is it pronounced "rooter" or "rowter"? Cheers, Andy From bpurvy at gmail.com Wed Nov 24 15:42:12 2021 From: bpurvy at gmail.com (Bob Purvy) Date: Wed, 24 Nov 2021 15:42:12 -0800 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: I've never heard anyone in the biz say "rooter." Have you? On Wed, Nov 24, 2021 at 3:25 PM Andrew G. Malis via Internet-history < internet-history at elists.isoc.org> wrote: > Is it pronounced "rooter" or "rowter"? > > Cheers, > Andy > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From jack at 3kitty.org Wed Nov 24 15:55:37 2021 From: jack at 3kitty.org (Jack Haverty) Date: Wed, 24 Nov 2021 15:55:37 -0800 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: <961baf23-f7f4-b7c6-6e7d-ada1a16bbd30@3kitty.org> I've heard, and used, both pronounciations.? Possibly it depends on whether you're using the term thinking about it as a noun or as a verb? Related question - what are these devices (routers, switches, bridges, ...) called in non-English languages??? Have the English words just been adopted, or have they been translated into the other language's equivalent? Jack On 11/24/21 3:42 PM, Bob Purvy via Internet-history wrote: > I've never heard anyone in the biz say "rooter." Have you? > > On Wed, Nov 24, 2021 at 3:25 PM Andrew G. Malis via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Is it pronounced "rooter" or "rowter"? >> >> Cheers, >> Andy >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> From tte at cs.fau.de Wed Nov 24 16:03:38 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Thu, 25 Nov 2021 01:03:38 +0100 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: On Wed, Nov 24, 2021 at 06:25:01PM -0500, Andrew G. Malis via Internet-history wrote: > Is it pronounced "rooter" or "rowter"? "rooter" = brits, "rowter" = colonists ? From olejacobsen at me.com Wed Nov 24 16:09:43 2021 From: olejacobsen at me.com (Ole Jacobsen) Date: Wed, 24 Nov 2021 16:09:43 -0800 Subject: [ih] For our next argument .... Message-ID: <90B1EA2B-517A-40B9-91F5-6DF0C3D44A9D@me.com> ?Yes, the entire proper English (British) speaking world says ?rooter?. Ole Ole J. Jacobsen Editor & Publisher The Internet Protocol Journal Office: +1 415 550-9433 Cell: +1 415 370-4628 T-Mobile: +1 415 889-9821 Docomo: +81 90 3337 9311? http://protocoljournal.org Sent from my iPhone > On 24 Nov 2021, at 15:42, Bob Purvy via Internet-history wrote: > ?I've never heard anyone in the biz say "rooter." Have you? > > On Wed, Nov 24, 2021 at 3:25 PM Andrew G. Malis via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Is it pronounced "rooter" or "rowter"? >> >> Cheers, >> Andy >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From jtk at dataplane.org Wed Nov 24 16:14:04 2021 From: jtk at dataplane.org (John Kristoff) Date: Wed, 24 Nov 2021 18:14:04 -0600 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: <20211124181404.36237253@p50.localdomain> On Thu, 25 Nov 2021 01:03:38 +0100 Toerless Eckert via Internet-history wrote: > "rooter" = brits, "rowter" = colonists ? Rich Seifert, a colonist, was clearly on the rooter side. His standard joke went something like... "I have both types at home. For a weekend projects one of them I use int the confines of my garage, which makes a heck of a racket and a mess all over the floor. The other one carves wood." John From jack at 3kitty.org Wed Nov 24 16:14:04 2021 From: jack at 3kitty.org (Jack Haverty) Date: Wed, 24 Nov 2021 16:14:04 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: Message-ID: <4a268300-5a0f-7912-81f1-411086d56ce3@3kitty.org> On 11/23/21 6:25 PM, John Shoch via Internet-history wrote: > That box (designed and built by BBN), no matter what > you call it, was not in 1969 the first to send bits over the Internet. I agree, but there are lots of people who think the Internet began with the ARPANET.?? I've argued, unsuccessfully, wih a few.? Terms mean what you think they mean... Jack From dhc at dcrocker.net Wed Nov 24 16:27:17 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Wed, 24 Nov 2021 16:27:17 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <4a268300-5a0f-7912-81f1-411086d56ce3@3kitty.org> References: <4a268300-5a0f-7912-81f1-411086d56ce3@3kitty.org> Message-ID: On 11/24/2021 4:14 PM, Jack Haverty via Internet-history wrote: > On 11/23/21 6:25 PM, John Shoch via Internet-history wrote: >> That box (designed and built by BBN), no matter what >> you call it, was not in 1969 the first to send bits over the Internet. > > I agree, but there are lots of people who think the Internet began with > the ARPANET.?? I've argued, unsuccessfully, wih a few.? Terms mean what > you think they mean... The Arpanet began a continuous, operational service that we use today. All sorts of components have been changed, over the decades, but the fact of daily, production-level interoperation among participating systems has been continuous, for roughly 50 years. Email has been in continuous operational service almost as long. In this case, the current, basic message looks almost the same as almost the earliest ,essages. But again, the service has operated without interruption, in spite of all its components being changed, too. We seem to be comfortable with the concept of continuity for telephone service and television service, in spite of all their technical changes. The distinction of the underlying technologies is fundamental, for technologists. But not for users. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From bpurvy at gmail.com Wed Nov 24 17:03:50 2021 From: bpurvy at gmail.com (Bob Purvy) Date: Wed, 24 Nov 2021 17:03:50 -0800 Subject: [ih] For our next argument .... In-Reply-To: <20211124181404.36237253@p50.localdomain> References: <20211124181404.36237253@p50.localdomain> Message-ID: It's "rooter" in the UK? Excuse me, Dr. Fauci says I have to go get my Borchestershire shot. On Wed, Nov 24, 2021 at 4:14 PM John Kristoff via Internet-history < internet-history at elists.isoc.org> wrote: > On Thu, 25 Nov 2021 01:03:38 +0100 > Toerless Eckert via Internet-history > wrote: > > > "rooter" = brits, "rowter" = colonists ? > > Rich Seifert, a colonist, was clearly on the rooter side. His standard > joke went something like... "I have both types at home. For a > weekend projects one of them I use int the confines of my garage, > which makes a heck of a racket and a mess all over the floor. The > other one carves wood." > > John > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From brian.e.carpenter at gmail.com Wed Nov 24 17:42:45 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Thu, 25 Nov 2021 14:42:45 +1300 Subject: [ih] multi-protocol routers, bridges (Was: "The First Router" on Jeopardy) In-Reply-To: <20211124220636.03A4B18C077@mercury.lcs.mit.edu> References: <20211124220636.03A4B18C077@mercury.lcs.mit.edu> Message-ID: <72371949-326f-87de-28bd-53bdde1873f2@gmail.com> > but I agree on the > 'multi-protocol was more common than single rotocol'. Yes. That's why at CERN, we built our own Ethernet bridges (called FRIGATEs) that bridged Ethernets over the homebrew CERNET backbone, up and running by 1986. https://cds.cern.ch/record/174243 https://cds.cern.ch/record/177513 https://cds.cern.ch/record/186966 On a trip to the US in ~1987, I visited Proteon, mainly to talk about Pronet-80, but also met Noel, who introduced me to the concept of multiprotocol routers. Regards Brian Carpenter On 25-Nov-21 11:06, Noel Chiappa via Internet-history wrote: > > From: Clem Cole > > > IP had not yet 'won' and a lot of site had multiple protocols running > > on their LANs. > > MIT had that problem in spades, which was exactly the genesis of the > multi-protocol router at MIT. > > CHAOS had a large lead in deployment at MIT, since it got rolling before > TCP/IP really did; I think the AI Lab guys got word of what was happening at > PARC, and decided to build a rough copy (hardware and, less identically, > protcol - the latter eventually ran over DIX Ethernet too). Once LCS got > rolling with the rings and TCP/IP, it clearly made sense to have all the LANs > carry both (or we'd have to rub both kinds of wire everwhere; much more cost > effective to carry both). After Dave Clark and I chatted with Dave Moon for a > while, we came with MUPPETs (MIT Universal Packets), to allow both CHAOS and > TCP/IP packets on the MIT LANs to share a common carriage format (and thus > only one kind of router). > > The CHAOS guys didn't show much movement towards actually implementhing them, > though (no surprise, not much incentive). So then it looked like carrying > each protocol independently would be easier to accomplish. That still left > the problem of inter-LAN gateways (routers), though - did those have to be > replicated? I came up with the multi-protocol router as a way to economize on > them. Dave Clark wasn't totally sold on the idea to begin with (he was in > charge of prodcing an MIT-wide campus LAN plan at the time, and included the > 'multi-protocol spine' as a possible solution, but without saying 'the only > plausible solution'). > > So I set off to write one, to show that it was viable. Somewhere in there the > Xerox grant showed up, so we had Experimental Ethernet and PUP to deal with > too. (My first kludgy router, written in MACRO-11, handled IP packets over > the rings, the Experimental Ethernet and CHAOSNET too, but IP-only, IIRC - I > ran across the code the other day, I could look.) > > (The AI guys actually implemented a service gateway to get to the Dover > printer; the AI-CHAOS-11 spoke CHAOS protocol to CHAOSNET hosts, and EFTP to > the Dover spooler. They had CHAOS protocol to NCP - later TCP - gateways too.) > > Something similar for Bill Yeager at Stanford, who independently re-invented > the multi-protocol spine/router idea; PUP and IP for them. > > > > Some were routable, some were not. > > Interesting point, but not true of MIT (and probably Stanford too); all our > protocols were router-able. That may be part of the latter attraction of > bridges, though. > > Note that bridges only _really_ work well on LANs with large, unique > interface network addresses; only after DIX Ethernet was that true. Early > LANs (Proteon rings, Omninet, etc, etc) had small addresses, and had to use > routers. > > > But mixed protocols was probably more the norm in commercials and I bet > > University circles than not. > > I'm not sure. It probably depended on what kind of boxes any particular place > had; not sure there was a commercial/academic difference, but I agree on the > 'multi-protocol was more common than single rotocol'. > > > > I know an early 3Com brouter but my memory is there were others. I > > think DEC made one or two. Noel can tell us what Proeton did. > > I know that at the start Proteon totally missed the bridge boat. I have to > hold up my hand for a lot of that; I was so focused on my 'a bridge looks > like a wire, but isn't' mantra (e.g. I thought we'd have to deal with > congestion explicitly - this was before Van Jacobsen), I missed out on the > advantages of bridges: trivial to install (especially with 'unique at > manufacture' interface network addresses), they supported non-routable > protocols, etc. > > Also, with Proteon's own LANs not having unique interface network addresses, we > weren't well-positioned to see the ease of use, etc, points. > > I think they did eventually add bridging functionality towards the end, but I > was not very involved there by then. > > Noel > From galmes at tamu.edu Wed Nov 24 17:47:38 2021 From: galmes at tamu.edu (Guy Almes) Date: Wed, 24 Nov 2021 20:47:38 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <315942109.667960.1637767799111@mail.yahoo.com> References: <315942109.667960.1637767799111@mail.yahoo.com> Message-ID: <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> Alex, It is true that the CIX episode marked a trend in this direction, but there are two reasons why your specific suggestion is not right. First, by 1987, when both the ARPAnet and the proto-NSFnet backbone were both operational, networks that connected to both had to decide which to use, and that led to interesting routing decisions. Problems encountered then led, for example, to creation of BGP. Similarly, again during the 1980s, a network might also connect to the NASA Science Internet and, again, have to make interesting routing choices. Second, recall that the initial CIX "exchange" was a single router with a T1 circuit from each of PSInet, Alternet, and CERFnet. I don't know exactly how that router was configured, but note that it was that CIX router that had to make many of the interesting routing decisions. Also, I think that the term 'router' as a Level-3 gateway was in heavy use prior to 1991. -- Guy On 11/24/21 10:29 AM, Alex McKenzie via Internet-history wrote: > I've stayed out of the discussion but listened with interest.? In fact, I have no clear recollection of when the terminology switched.? But I suspect it happened around the time the first CIX (Commercial Internet Exchange) agreement was reached.? Before that there was a "backbone" network (first ARPAnet and then NSFnet).? Packets went from your network to the backbone, across the backbone, and then to the destination network.? But after the CIX agreement there became a meaningful choice of routes, and the gateways had to start figuring out what the set of available routes was and choose one.? I understand this is an oversimplification of the internet structure, but I think it is a meaningful one. > Cheers,Alex > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://urldefense.com/v3/__https://elists.isoc.org/mailman/listinfo/internet-history__;!!KwNVnqRv!QIWtjydlEg0JAfsQNrajSaK1_e6vZT1uICYHfASZKIT-OFn4axRJrCuqS8Tc_w$ > > . From alejandroacostaalamo at gmail.com Wed Nov 24 18:01:07 2021 From: alejandroacostaalamo at gmail.com (Alejandro Acosta) Date: Wed, 24 Nov 2021 22:01:07 -0400 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: <9109407c-e7a3-2518-1fdc-af7f5fd181b8@gmail.com> Yes, actually many people says rooter. I prefer to pronounce "rowter" but I have to admit? that I have said "rooter" and so far so good. Alejandro, P.S. I'm not a native English speaker On 24/11/21 7:42 PM, Bob Purvy via Internet-history wrote: > I've never heard anyone in the biz say "rooter." Have you? > > On Wed, Nov 24, 2021 at 3:25 PM Andrew G. Malis via Internet-history < > internet-history at elists.isoc.org> wrote: > >> Is it pronounced "rooter" or "rowter"? >> >> Cheers, >> Andy >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history >> From brian.e.carpenter at gmail.com Wed Nov 24 18:06:40 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Thu, 25 Nov 2021 15:06:40 +1300 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: <67eb5257-3e6d-898f-976d-8d153391bb79@gmail.com> On 25-Nov-21 13:03, Toerless Eckert via Internet-history wrote: > On Wed, Nov 24, 2021 at 06:25:01PM -0500, Andrew G. Malis via Internet-history wrote: >> Is it pronounced "rooter" or "rowter"? > > "rooter" = brits, "rowter" = colonists ? It's a bit more complicated. Firstly, I think you'll sometimes find the "British" pronunciation on the East Coast of the USA, or even occcasionally in the Midwest. Secondly, when I first taught networking in NZ, where the accent is much closer to British English than US (and very different from Australian), I was *strongly* advised to say "rowt" and "rowter". Why? Because in NZ and Australia, "root" as a transitive verb has a slang connotation that is entirely amusing to a younger audience. So for colonists down under, things are different. Mick Jagger confuses things even more: https://www.youtube.com/watch?v=Q131ZJ6YkG0 Brian From tte at cs.fau.de Wed Nov 24 18:53:18 2021 From: tte at cs.fau.de (Toerless Eckert) Date: Thu, 25 Nov 2021 03:53:18 +0100 Subject: [ih] For our next argument .... In-Reply-To: <67eb5257-3e6d-898f-976d-8d153391bb79@gmail.com> References: <67eb5257-3e6d-898f-976d-8d153391bb79@gmail.com> Message-ID: Thanks, Brian I think i exhausted my ability to intake details on the matter in 1999 when i worked in London and everybody around me came from randomn places around the world, all parts of the prior british empire included, and we discussed the matter fewerishly at lunch. I just wanted to go for the cheap pun ;-) Cheers Toerless On Thu, Nov 25, 2021 at 03:06:40PM +1300, Brian E Carpenter wrote: > On 25-Nov-21 13:03, Toerless Eckert via Internet-history wrote: > > On Wed, Nov 24, 2021 at 06:25:01PM -0500, Andrew G. Malis via Internet-history wrote: > > > Is it pronounced "rooter" or "rowter"? > > > > "rooter" = brits, "rowter" = colonists ? > > It's a bit more complicated. Firstly, I think you'll sometimes find the "British" pronunciation on the East Coast of the USA, or even occcasionally in the Midwest. > > Secondly, when I first taught networking in NZ, where the accent is much closer to British English than US (and very different from Australian), I was *strongly* advised to say "rowt" and "rowter". Why? Because in NZ and Australia, "root" as a transitive verb has a slang connotation that is entirely amusing to a younger audience. So for colonists down under, things are different. > > Mick Jagger confuses things even more: https://www.youtube.com/watch?v=Q131ZJ6YkG0 > > Brian -- --- tte at cs.fau.de From johnl at iecc.com Wed Nov 24 19:22:00 2021 From: johnl at iecc.com (John Levine) Date: 24 Nov 2021 22:22:00 -0500 Subject: [ih] For our next argument .... In-Reply-To: Message-ID: <20211125032201.553F2308B89E@ary.local> It appears that Andrew G. Malis via Internet-history said: >Is it pronounced "rooter" or "rowter"? Yes. Thanks for asking. R's, John From wayne at playaholic.com Wed Nov 24 20:14:58 2021 From: wayne at playaholic.com (Wayne Hathaway) Date: Wed, 24 Nov 2021 23:14:58 -0500 Subject: [ih] For our next argument .... Message-ID: <1637813698.7vw3pyou4gssok00@hostingemail.digitalspace.net> I remember a similar reasoning, that one pronunciation got confused with the Unix "root" so everybody should use the other. At least that's how I was indoctrinated once I found myself in the Unix world. wayne On Thu, 25 Nov 2021 15:06:40 +1300, Brian E Carpenter via Internet-history wrote: >> On 25-Nov-21 13:03, Toerless Eckert via Internet-history wrote: >> > On Wed, Nov 24, 2021 at 06:25:01PM -0500, Andrew G. Malis via Internet-history wrote: >> >> Is it pronounced "rooter" or "rowter"? >> > >> > "rooter" = brits, "rowter" = colonists ? >> >> It's a bit more complicated. Firstly, I think you'll sometimes find the "British" pronunciation on the East Coast of the USA, or even occcasionally in the Midwest. >> >> Secondly, when I first taught networking in NZ, where the accent is much closer to British English than US (and very different from Australian), I was *strongly* advised to say "rowt" and "rowter". Why? Because in NZ and Australia, "root" as a transitive verb has a slang connotation that is entirely amusing to a younger audience. So for colonists down under, things are different. >> >> Mick Jagger confuses things even more: https://www.youtube.com/watch?v=Q131ZJ6YkG0 >> >> Brian >> -- >> Internet-history mailing list >> Internet-history at elists.isoc.org >> https://elists.isoc.org/mailman/listinfo/internet-history From mark at good-stuff.co.uk Thu Nov 25 04:26:01 2021 From: mark at good-stuff.co.uk (Mark Goodge) Date: Thu, 25 Nov 2021 12:26:01 +0000 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: <813ec8be-1511-ef87-fade-db744cbcfe8f@good-stuff.co.uk> On 24/11/2021 23:25, Andrew G. Malis via Internet-history wrote: > Is it pronounced "rooter" or "rowter"? It depends where you are. In British English, it's "rooter". In US English, it's "rowter". That's just a linguistic difference, not a technical one. British English pronounces "route" as "root", and hence "router", meaning something which routes, follows along the same principle. In British English, "rowter" means the woodworking tool, also spelled "router" but deriving from the verb "rout" (no e) so having a different etymology. Mark From jmamodio at gmail.com Thu Nov 25 06:37:09 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Thu, 25 Nov 2021 08:37:09 -0600 Subject: [ih] For our next argument .... In-Reply-To: References: Message-ID: In Spanish, ruter. Happy Thanks Giving y?all -Jorge > On Nov 24, 2021, at 5:25 PM, Andrew G. Malis via Internet-history wrote: > > ?Is it pronounced "rooter" or "rowter"? > > Cheers, > Andy > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From amckenzie3 at yahoo.com Thu Nov 25 09:24:17 2021 From: amckenzie3 at yahoo.com (Alex McKenzie) Date: Thu, 25 Nov 2021 17:24:17 +0000 (UTC) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> Message-ID: <432011064.2634901.1637861057847@mail.yahoo.com> Guy, I agree that my guess about the influence of the CIX is the wrong time frame.? I should have done a bit of research before speaking.? Thanks for the correction. Cheers,Alex On Wednesday, November 24, 2021, 08:47:43 PM EST, Guy Almes wrote: Alex, ? It is true that the CIX episode marked a trend in this direction, but there are two reasons why your specific suggestion is not right. ? First, by 1987, when both the ARPAnet and the proto-NSFnet backbone were both operational, networks that connected to both had to decide which to use, and that led to interesting routing decisions.? Problems encountered then led, for example, to creation of BGP. ? Similarly, again during the 1980s, a network might also connect to the NASA Science Internet and, again, have to make interesting routing choices. ? Second, recall that the initial CIX "exchange" was a single router with a T1 circuit from each of PSInet, Alternet, and CERFnet.? I don't know exactly how that router was configured, but note that it was that CIX router that had to make many of the interesting routing decisions. ? Also, I think that the term 'router' as a Level-3 gateway was in heavy use prior to 1991. ??? -- Guy On 11/24/21 10:29 AM, Alex McKenzie via Internet-history wrote: > I've stayed out of the discussion but listened with interest.? In fact, I have no clear recollection of when the terminology switched.? But I suspect it happened around the time the first CIX (Commercial Internet Exchange) agreement was reached.? Before that there was a "backbone" network (first ARPAnet and then NSFnet).? Packets went from your network to the backbone, across the backbone, and then to the destination network.? But after the CIX agreement there became a meaningful choice of routes, and the gateways had to start figuring out what the set of available routes was and choose one.? I understand this is an oversimplification of the internet structure, but I think it is a meaningful one. > Cheers,Alex > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://urldefense.com/v3/__https://elists.isoc.org/mailman/listinfo/internet-history__;!!KwNVnqRv!QIWtjydlEg0JAfsQNrajSaK1_e6vZT1uICYHfASZKIT-OFn4axRJrCuqS8Tc_w$ > > . From gregskinner0 at icloud.com Thu Nov 25 09:46:44 2021 From: gregskinner0 at icloud.com (Greg Skinner) Date: Thu, 25 Nov 2021 09:46:44 -0800 Subject: [ih] For our next argument .... and a proposal to make legacy IPv4 address space available In-Reply-To: <961baf23-f7f4-b7c6-6e7d-ada1a16bbd30@3kitty.org> References: <961baf23-f7f4-b7c6-6e7d-ada1a16bbd30@3kitty.org> Message-ID: For examples in Spanish, take a look at the Spanish-language version of the Wikipedia ?Router? page Some of the content looks as if it was translated directly from the English version. It contains references to the involvement of Xerox PARC, etc. Incidentally, if you?ve been following the latest discussions about making legacy IPv4 address space available, it?s now being discussed on the LACNOG mailing list (subject: Draft: Unicast Use of the Formerly Reserved 127/8 ). You?ll see how the English words and phrases for networking devices, etc. are used in common discourse among speakers of languages spoken in Latin American countries. ?gregbo > On Nov 24, 2021, at 3:55 PM, Jack Haverty wrote: > > I've heard, and used, both pronounciations.? Possibly it depends on > whether you're using the term thinking about it as a noun or as a verb? > > Related question - what are these devices (routers, switches, bridges, > ...) called in non-English languages??? Have the English words just been > adopted, or have they been translated into the other language's equivalent? > > Jack > > > On 11/24/21 3:42 PM, Bob Purvy via Internet-history wrote: >> I've never heard anyone in the biz say "rooter." Have you? >> >> On Wed, Nov 24, 2021 at 3:25 PM Andrew G. Malis via Internet-history < >> internet-history at elists.isoc.org> wrote: >> >>> Is it pronounced "rooter" or "rowter"? >>> >>> Cheers, >>> Andy >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history >>> > > > From galmes at tamu.edu Thu Nov 25 10:43:01 2021 From: galmes at tamu.edu (Guy Almes) Date: Thu, 25 Nov 2021 13:43:01 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <432011064.2634901.1637861057847@mail.yahoo.com> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> <432011064.2634901.1637861057847@mail.yahoo.com> Message-ID: <5a5a80f5-5ab2-f30e-1ab6-cc022434e729@tamu.edu> Alex, Not a problem. Your post actually made me curious about how the CIX router was configured. Who decided which network to point to for a prefix connected to two or more CIX members? And, of course, it didn't take long for exchange points to shift to level-2 networks such as the FIX (federal) exchanges at Univ Maryland and at (I think) NASA Ames. And the cleverly named MAE-East came soon thereafter. -- Guy On 11/25/21 12:24 PM, Alex McKenzie wrote: > Guy, > > I agree that my guess about the influence of the CIX is the wrong time > frame.? I should have done a bit of research before speaking.? Thanks > for the correction. > > Cheers, > Alex > > On Wednesday, November 24, 2021, 08:47:43 PM EST, Guy Almes > wrote: > > > Alex, > ? It is true that the CIX episode marked a trend in this direction, but > there are two reasons why your specific suggestion is not right. > > ? First, by 1987, when both the ARPAnet and the proto-NSFnet backbone > were both operational, networks that connected to both had to decide > which to use, and that led to interesting routing decisions.? Problems > encountered then led, for example, to creation of BGP. > ? Similarly, again during the 1980s, a network might also connect to > the NASA Science Internet and, again, have to make interesting routing > choices. > > ? Second, recall that the initial CIX "exchange" was a single router > with a T1 circuit from each of PSInet, Alternet, and CERFnet.? I don't > know exactly how that router was configured, but note that it was that > CIX router that had to make many of the interesting routing decisions. > > ? Also, I think that the term 'router' as a Level-3 gateway was in > heavy use prior to 1991. > ??? -- Guy > > On 11/24/21 10:29 AM, Alex McKenzie via Internet-history wrote: > > I've stayed out of the discussion but listened with interest.? In > fact, I have no clear recollection of when the terminology switched. > But I suspect it happened around the time the first CIX (Commercial > Internet Exchange) agreement was reached.? Before that there was a > "backbone" network (first ARPAnet and then NSFnet).? Packets went from > your network to the backbone, across the backbone, and then to the > destination network.? But after the CIX agreement there became a > meaningful choice of routes, and the gateways had to start figuring out > what the set of available routes was and choose one.? I understand this > is an oversimplification of the internet structure, but I think it is a > meaningful one. > > Cheers,Alex > > > -- > > Internet-history mailing list > > Internet-history at elists.isoc.org > > > > https://urldefense.com/v3/__https://elists.isoc.org/mailman/listinfo/internet-history__;!!KwNVnqRv!QIWtjydlEg0JAfsQNrajSaK1_e6vZT1uICYHfASZKIT-OFn4axRJrCuqS8Tc_w$ > > > > > > . From jmamodio at gmail.com Thu Nov 25 11:10:08 2021 From: jmamodio at gmail.com (Jorge Amodio) Date: Thu, 25 Nov 2021 13:10:08 -0600 Subject: [ih] For our next argument .... and a proposal to make legacy IPv4 address space available In-Reply-To: References: <961baf23-f7f4-b7c6-6e7d-ada1a16bbd30@3kitty.org> Message-ID: When we discuss technical matters, not just Internet, in Spanish speaking circles, it is often more easy and clear for everybody to use the English word for some stuff. It is common that even being "Spanish" translations vary from country to country, for example in South America and other countries Computer is Computadora, in Spain is Ordenador. Many many years ago there was a dim effort to translate some key RFC's to Spanish, it ended being more easy to teach English to those in need to read them :-) Cheers Jorge On Thu, Nov 25, 2021 at 11:46 AM Greg Skinner via Internet-history < internet-history at elists.isoc.org> wrote: > For examples in Spanish, take a look at the Spanish-language version of > the Wikipedia ?Router? page Some > of the content looks as if it was translated directly from the English > version. It contains references to the involvement of Xerox PARC, etc. > > Incidentally, if you?ve been following the latest discussions about making > legacy IPv4 address space available, it?s now being discussed on the LACNOG > mailing list (subject: Draft: Unicast Use of the Formerly Reserved 127/8 < > https://mail.lacnic.net/pipermail/lacnog/2021-November/thread.html>). > You?ll see how the English words and phrases for networking devices, etc. > are used in common discourse among speakers of languages spoken in Latin > American countries. > > ?gregbo > > > On Nov 24, 2021, at 3:55 PM, Jack Haverty wrote: > > > > I've heard, and used, both pronounciations.? Possibly it depends on > > whether you're using the term thinking about it as a noun or as a verb? > > > > Related question - what are these devices (routers, switches, bridges, > > ...) called in non-English languages??? Have the English words just been > > adopted, or have they been translated into the other language's > equivalent? > > > > Jack > > > > > > On 11/24/21 3:42 PM, Bob Purvy via Internet-history wrote: > >> I've never heard anyone in the biz say "rooter." Have you? > >> > >> On Wed, Nov 24, 2021 at 3:25 PM Andrew G. Malis via Internet-history < > >> internet-history at elists.isoc.org> wrote: > >> > >>> Is it pronounced "rooter" or "rowter"? > >>> > >>> Cheers, > >>> Andy > >>> -- > >>> Internet-history mailing list > >>> Internet-history at elists.isoc.org > >>> https://elists.isoc.org/mailman/listinfo/internet-history > >>> > > > > > > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history > From jack at 3kitty.org Thu Nov 25 12:16:50 2021 From: jack at 3kitty.org (Jack Haverty) Date: Thu, 25 Nov 2021 12:16:50 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> Message-ID: On 11/24/21 5:47 PM, Guy Almes via Internet-history wrote: > First, by 1987, when both the ARPAnet and the proto-NSFnet backbone > were both operational, networks that connected to both had to decide > which to use, and that led to interesting routing decisions. Problems > encountered then led, for example, to creation of BGP. FYI, for The Historians, I offer a little of the earlier history, all from the 1980-1983 timeframe. At the ICCB (before it was renamed IAB) meetings, there was a list kept on the whiteboard of "things that need to get figured out". One of the items on that list was "Routing", which included issues like "Type Of Service", "Shortest Path First", and "Policy Routing" -- all part of the "How should routing behave in the future Internet?" topic.? There were two concrete motivators for these issues. The Internet had started to evolve from its "fuzzy peach" stage, where essentially the ARPANET was the peach surrounded by a fuzzy layer of? LANs, into a richer topology where there were actually choices to be made. First, SATNET had linked the US and Europe for a while, but Bob Kahn initiated the addition of a transatlantic path using the public X.25 service.?? The interconnect was called a "VAN Gateway" (VAN stood for Value Added Network but I never did understand what that really meant). ? The VAN gateway essentially added an interface option to the existing gateways, allowing them to connect to the public X.25 networks, and use them as a "circuit" (albeit virtual) between gateways.? In effect, the entire X.25 public network system was made an available component of the Internet; it became just another network that could be used within the overall Internet. The "dial up" nature of the X.25 service also introduced the possibility of dynamic configuration of the Internet topology -- e.g., adding or deleting circuits between pairs of gateways as a situation warranted, simply by using the dialup infrastructure of the public X.27/X.75 system.? We called that something like "Dynamic Adaptive Topology", but I don't recall ever actually trying to use that capability except on the single US<->EU path in parallel with SATNET. This "VAN" capability was used to create a topology where there were two ways IP datagrams could cross the Atlantic.?? Bringing economics into the picture, the SATNET path was funded by ARPA, to be used only for ARPA-approved projects.?? The X.25 path was funded by whoever opened the circuit first (which we, as ARPA contractors, silently engineered to be usually the European side; seemed like the right thing to do).?? This issue appeared on the ICCB's to-do list as "Policy Based Routing". Pending the "real" solution, I don't recall exactly how the gateways back then made the choice of path for each packet; my vague recollection is that it had something to do with destination addresses - i.e., a host might have two distinct IP addresses, one for use via SATNET and the other for use via X.25.?? And a single physical net would be assigned two network numbers (no shortage then), one for use via X.25 and the other for use via SATNET.? IIRC, that's how the UCL network in London was configured.? The early Internet had to sometimes use patching plaster and baling wire to keep it going as the research sought the right way for the future. Second, the Wideband Net, a satellite-based network spanning the US, was made part of the Internet topology by gateways between it and the ARPANET.?? There were then multiple network paths across the US.?? But the gateways' routing metric of "hops" would never cause any datagrams to be sent over the Wideband Net.?? Since the Wideband Net was only interconnected by gateways to ARPANET nodes, any route that used the Wideband Net would necessarily be 2 hops longer than a direct path across the ARPANET.?? The routing mechanisms would never make such a choice.??? This issue was captured on the ICCB's list as "Expressway Routing", a reference to the need for car drivers to make a decision to head toward the nearest freeway entrance, rather than taking a direct route toward their destination, in order to get to their destination faster. I don't recall how people experimented with the Wideband Net, i.e., how they got datagrams to actually flow over it.?? Perhaps that was a use of the "Source Routing" mechanisms in the IP headers.?? Maybe someone else remembers.... We didn't know how to best address these situations, but of course there were a lot of ideas.? In addition, the existing Internet lacked some basic mechanisms that seemed to be necessary.? In particular, the use of "hops" to determine which path was the shortest was woefully inadequate.?? A "hop" through a Satellite net might be expected to take much longer than a hop through a terrestrial net, simply due to Physics. ? But a hop through the ARPANET traversing many IMPs, when the net was congested, might actually take longer than a satellite transit.? A time-based metric was not feasible in the gateways without some means of accurately measuring time, at a precision of milliseconds, in the routers scattered across the continents. Dave Mills was on the ICCB, and he took on this quest with unbridled energy and determination.?? NTP was the result - an impressive piece of engineering.?? Using NTP, computers (e.g., routers, gateways, hosts, servers, whatever you call them) can maintain highly synchronized clocks, and measure actual transit times of IP datagrams for use in calculating "shortest path".??? Everyone can thank Dave and his crew that your computers know what time it is today. The Time mechanisms would be helpful, but much more was needed to handle "Policy Based" and "Expressway" routing situations.? Lots of people had ideas and wanted them put into the "core gateways" that BBN operated.?? But doing that kind of experimentation and also keeping the Internet core reliably running 24x7 was a struggle. I was also on the ICCB at the time, and I recruited Dr. Eric Rosen back at BBN to help think about it.?? He and I had numerous day-long sessions with a whiteboard.? The result was EGP - the Exterior Gateway Protocol.?? If you read Eric's now ancient RFC defining EGP, you'll see that it was not intended as a routing protocol.?? Rather it was more of a "firewall" mechanism that would allow the Internet to be carved up into pieces, each of which was implemented and operated at arm's length from the others but could interoperate to present a single Internet to the end users. The intent was that such a mechanism would make it possible for some collection of gateways (e.g., the "core gateways" of the Internet at that time) to be operated as a reliable service, while also enabling lots of other collections of gateways to be used as guinea pigs for all sorts of experiments to try out various ideas that had come up.?? Each such collection was called an "Autonomous System" of gateways using some particular technical mechanisms and under some single operator's control.?? EGP was a mechanism to permit reliable operational services to coexist in the Internet with research and experimentation. When the ideas had been tried, and the traditional "rough consensus" emerged to identify the best system design, the new algorithms, mechanisms, protocols, and anything else needed, would be instantiated in a new Autonomous System, which would then grow as the new system was deployed - much as the ARPANET has served as the nursery for the fledgling Internet, with all IMPs disappearing over time as they were replaced by routers directly connected with wires. That's where my direct involvement in the "research" stopped, as I went more into deploying and operating network stuff, from about mid-1983 on.?? Perhaps someone else can fill in more gaps in the History. Enjoy, Jack Haverty From gregskinner0 at icloud.com Thu Nov 25 15:23:27 2021 From: gregskinner0 at icloud.com (Greg Skinner) Date: Thu, 25 Nov 2021 15:23:27 -0800 Subject: [ih] For our next argument .... and a proposal to make legacy IPv4 address space available In-Reply-To: References: <961baf23-f7f4-b7c6-6e7d-ada1a16bbd30@3kitty.org> Message-ID: <23C049B9-4067-491F-B30F-7EA1AE32AA3D@icloud.com> OK, fair enough. I didn?t mean to imply that the thread was indicative of how Spanish speakers communicate about technical matters amongst themselves. It was an example that I thought people here would recognize, especially since the issue came up here some time ago. I occasionally read the LACNIC mailing list and watch some of the video presentations. There?s a variety of discourse ? some English only, some Spanish only, some mixed. ?gregbo > On Nov 25, 2021, at 11:10 AM, Jorge Amodio wrote: > > > When we discuss technical matters, not just Internet, in Spanish speaking circles, it is often more easy and clear for everybody to use the English word for some stuff. > > It is common that even being "Spanish" translations vary from country to country, for example in South America and other countries Computer is Computadora, in Spain is Ordenador. > > Many many years ago there was a dim effort to translate some key RFC's to Spanish, it ended being more easy to teach English to those in need to read them :-) > > Cheers > Jorge > > > On Thu, Nov 25, 2021 at 11:46 AM Greg Skinner via Internet-history > wrote: > For examples in Spanish, take a look at the Spanish-language version of the Wikipedia ?Router? page > Some of the content looks as if it was translated directly from the English version. It contains references to the involvement of Xerox PARC, etc. > > Incidentally, if you?ve been following the latest discussions about making legacy IPv4 address space available, it?s now being discussed on the LACNOG mailing list (subject: Draft: Unicast Use of the Formerly Reserved 127/8 >). You?ll see how the English words and phrases for networking devices, etc. are used in common discourse among speakers of languages spoken in Latin American countries. > > ?gregbo > > > On Nov 24, 2021, at 3:55 PM, Jack Haverty > wrote: > > > > I've heard, and used, both pronounciations.? Possibly it depends on > > whether you're using the term thinking about it as a noun or as a verb? > > > > Related question - what are these devices (routers, switches, bridges, > > ...) called in non-English languages??? Have the English words just been > > adopted, or have they been translated into the other language's equivalent? > > > > Jack > > > > > > On 11/24/21 3:42 PM, Bob Purvy via Internet-history wrote: > >> I've never heard anyone in the biz say "rooter." Have you? > >> > >> On Wed, Nov 24, 2021 at 3:25 PM Andrew G. Malis via Internet-history < > >> internet-history at elists.isoc.org > wrote: > >> > >>> Is it pronounced "rooter" or "rowter"? > >>> > >>> Cheers, > >>> Andy > >>> -- > >>> Internet-history mailing list > >>> Internet-history at elists.isoc.org > >>> https://elists.isoc.org/mailman/listinfo/internet-history > >>> > > > > > > > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From amckenzie3 at yahoo.com Fri Nov 26 09:28:26 2021 From: amckenzie3 at yahoo.com (Alex McKenzie) Date: Fri, 26 Nov 2021 17:28:26 +0000 (UTC) Subject: [ih] My 2013 paper on the History of the Internet References: <184824475.660661.1637947706565.ref@mail.yahoo.com> Message-ID: <184824475.660661.1637947706565@mail.yahoo.com> In 2014 a paper I was hired to write about the history of the Internet was published in the 3-volume "Discoveries in Modern Science." I was required to cover several specific items which the publishers considered critical, whether I thought so or not. Of course, with a restricted word limit the big question for an author is what to include and what to leave out.? I'm sure other authors would have made different choices from mine. I was recently granted permission to post a copy of this article on my website.? If anyone is interested it can be found athttp://alexmckenzie.weebly.com/history-of-the-internet.html Cheers,Alex From olejacobsen at me.com Fri Nov 26 09:31:52 2021 From: olejacobsen at me.com (Ole Jacobsen) Date: Fri, 26 Nov 2021 09:31:52 -0800 Subject: [ih] My 2013 paper on the History of the Internet In-Reply-To: <184824475.660661.1637947706565@mail.yahoo.com> References: <184824475.660661.1637947706565.ref@mail.yahoo.com> <184824475.660661.1637947706565@mail.yahoo.com> Message-ID: <43F9401A-3F28-473D-A1FA-A008289001FF@me.com> Just in case you're getting the same "there is no application for athttp" message, here is the correct link: http://alexmckenzie.weebly.com/history-of-the-internet.html > On Nov 26, 2021, at 09:28, Alex McKenzie via Internet-history wrote: > > In 2014 a paper I was hired to write about the history of the Internet was published in the 3-volume "Discoveries in Modern Science." I was required to cover several specific items which the publishers considered critical, whether I thought so or not. Of course, with a restricted word limit the big question for an author is what to include and what to leave out. I'm sure other authors would have made different choices from mine. > I was recently granted permission to post a copy of this article on my website. If anyone is interested it can be found athttp://alexmckenzie.weebly.com/history-of-the-internet.html > > Cheers,Alex > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history Ole J. Jacobsen Editor and Publisher The Internet Protocol Journal Office: +1 415-550-9433 Cell: +1 415-370-4628 Web: protocoljournal.org E-mail: olejacobsen at me.com E-mail: ole at protocoljournal.org Skype: organdemo From scott.brim at gmail.com Fri Nov 26 09:34:34 2021 From: scott.brim at gmail.com (Scott Brim) Date: Fri, 26 Nov 2021 12:34:34 -0500 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> Message-ID: I'm just going to toss something in here from 1651. The second half is particularly relevant. There is nothing new under the sun etc. By this it appears how necessary it is for any man that aspires to true knowledge, to examine the definitions of former authors; and either to correct them, where they are negligently set down, or to make them himself. For the errors of definitions multiply themselves according as the reckoning proceeds, and lead men into absurdities, which at least they see, but cannot avoid, without reckoning anew from the beginning. -- Hobbes, Leviathan From galmes at tamu.edu Fri Nov 26 11:31:06 2021 From: galmes at tamu.edu (Guy Almes) Date: Fri, 26 Nov 2021 14:31:06 -0500 Subject: [ih] My 2013 paper on the History of the Internet In-Reply-To: <184824475.660661.1637947706565@mail.yahoo.com> References: <184824475.660661.1637947706565.ref@mail.yahoo.com> <184824475.660661.1637947706565@mail.yahoo.com> Message-ID: <1d94e873-f16f-d0a2-48f8-ac22b6b56dbd@tamu.edu> Alex, A very nice paper. Most folks on the list could find a few bugs in it, and each of us, were we writing it, would shift some points of emphasis, but overall I found it very well written and balanced. Thanks for sharing it, -- Guy On 11/26/21 12:28 PM, Alex McKenzie via Internet-history wrote: > In 2014 a paper I was hired to write about the history of the Internet was published in the 3-volume "Discoveries in Modern Science." I was required to cover several specific items which the publishers considered critical, whether I thought so or not. Of course, with a restricted word limit the big question for an author is what to include and what to leave out.? I'm sure other authors would have made different choices from mine. > I was recently granted permission to post a copy of this article on my website.? If anyone is interested it can be found athttp://alexmckenzie.weebly.com/history-of-the-internet.html > > Cheers,Alex > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://urldefense.com/v3/__https://elists.isoc.org/mailman/listinfo/internet-history__;!!KwNVnqRv!URntTTY8CsbSombb12qF6yAhJB3j6S1qYKJjNkK3LXFECW5FdALsyQSwuVqqlA$ > From b_a_denny at yahoo.com Fri Nov 26 15:06:19 2021 From: b_a_denny at yahoo.com (Barbara Denny) Date: Fri, 26 Nov 2021 23:06:19 +0000 (UTC) Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> Message-ID: <1255978921.5754777.1637967979175@mail.yahoo.com> I went off to see what more I could find to check this timing out. Hope you don't mind me using myself in the search since that seemed easiest.? ?In the 1990s I took over a project involving the DARTnet testbed from Paul McKinney when he left SRI.? He had a paper at IEEE INFOCOM 1990 on SFQ (Stochastic Fairness Queuing). The conference was in June. In that paper,? he uses the term gateway. At the end of the project,? I submitted a technical report to the client on what I had done after he left SRI with SFQ. (In case you are interested,? I also did more work with an SFQ extension involving Virtual Clock. I had to use ST-II since RSVP didn't exist yet). This report was dated March 1993,? and I used the term router to describe the configuration.?? I was able to find a copy of my report online. https://ntrs.nasa.gov/api/citations/19970014215/downloads/19970014215.pdf I am sure you can find a copy of Paul's IEEE INFOCOM paper. barbara On Wednesday, November 24, 2021, 05:47:51 PM PST, Guy Almes via Internet-history wrote: Alex, ? It is true that the CIX episode marked a trend in this direction, but there are two reasons why your specific suggestion is not right. ? First, by 1987, when both the ARPAnet and the proto-NSFnet backbone were both operational, networks that connected to both had to decide which to use, and that led to interesting routing decisions.? Problems encountered then led, for example, to creation of BGP. ? Similarly, again during the 1980s, a network might also connect to the NASA Science Internet and, again, have to make interesting routing choices. ? Second, recall that the initial CIX "exchange" was a single router with a T1 circuit from each of PSInet, Alternet, and CERFnet.? I don't know exactly how that router was configured, but note that it was that CIX router that had to make many of the interesting routing decisions. ? Also, I think that the term 'router' as a Level-3 gateway was in heavy use prior to 1991. ??? -- Guy On 11/24/21 10:29 AM, Alex McKenzie via Internet-history wrote: > I've stayed out of the discussion but listened with interest.? In fact, I have no clear recollection of when the terminology switched.? But I suspect it happened around the time the first CIX (Commercial Internet Exchange) agreement was reached.? Before that there was a "backbone" network (first ARPAnet and then NSFnet).? Packets went from your network to the backbone, across the backbone, and then to the destination network.? But after the CIX agreement there became a meaningful choice of routes, and the gateways had to start figuring out what the set of available routes was and choose one.? I understand this is an oversimplification of the internet structure, but I think it is a meaningful one. > Cheers,Alex > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://urldefense.com/v3/__https://elists.isoc.org/mailman/listinfo/internet-history__;!!KwNVnqRv!QIWtjydlEg0JAfsQNrajSaK1_e6vZT1uICYHfASZKIT-OFn4axRJrCuqS8Tc_w$ > > . -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From geoff at iconia.com Fri Nov 26 16:59:26 2021 From: geoff at iconia.com (the keyboard of geoff goodfellow) Date: Fri, 26 Nov 2021 14:59:26 -1000 Subject: [ih] My 2013 paper on the History of the Internet In-Reply-To: <184824475.660661.1637947706565@mail.yahoo.com> References: <184824475.660661.1637947706565.ref@mail.yahoo.com> <184824475.660661.1637947706565@mail.yahoo.com> Message-ID: Alex, please pardon the always being the unrelenting curious and inquisitive mind here, but would you be willing to share what were the several specific items which the publishers considered critical and you were required to cover in your history of the Internet paper, pretty please? geoff On Fri, Nov 26, 2021 at 7:28 AM Alex McKenzie via Internet-history < internet-history at elists.isoc.org> wrote: > In 2014 a paper I was hired to write about the history of the Internet was > published in the 3-volume "Discoveries in Modern Science." I was required > to cover several specific items which the publishers considered critical, > whether I thought so or not. Of course, with a restricted word limit the > big question for an author is what to include and what to leave out. I'm > sure other authors would have made different choices from mine. > I was recently granted permission to post a copy of this article on my > website. If anyone is interested it can be found at http:// > alexmckenzie.weebly.com/history-of-the-internet.html > > Cheers,Alex > -- Geoff.Goodfellow at iconia.com living as The Truth is True From geoff at iconia.com Fri Nov 26 17:23:58 2021 From: geoff at iconia.com (the keyboard of geoff goodfellow) Date: Fri, 26 Nov 2021 15:23:58 -1000 Subject: [ih] ARPA Information Processing Techniques Office history (was: Alex McKenzie's 2013 paper on the History of the Internet) In-Reply-To: References: <184824475.660661.1637947706565.ref@mail.yahoo.com> <184824475.660661.1637947706565@mail.yahoo.com> Message-ID: in reading Alex's 2013 paper on the History of the Internet yours truly saw mention as to who was the first IPTO director (Lick) and then immediate mention that Robert W. ??Bob?? Taylor (1932? ) became the third IPTO director in 1966 -- but with no mention as to who was the second IPTO director... so over to Wikipedia it was and yours truly found at https://en.wikipedia.org/wiki/Information_Processing_Techniques_Office under "Later history Ivan Sutherland replaced J. C. R. Licklider as IPTO's director when Licklider left ARPA in 1964.[7][8] Sutherland was 26 years old at the time. Bob Taylor was hired as Sutherland's assistant in 1965 and became director in 1966.[9] During Taylor's tenure, the IPTO facility consisted of a spacious office for the director in Ring D of The Pentagon and a small "terminal room" with remote terminals to mainframe computers at MIT, the University of California, Berkeley and the AN/FSQ-32 in Santa Monica.[10] The staff at the Pentagon consisted of the director and his secretary.[11] The budget was $19 million which funded computer research projects at MIT and other institutions in Massachusetts and California.[12] In 1966 Taylor went to ARPA, on Ring E, for funding to create a computer network that used interactive computing. He got $1 million and hired Lawrence Roberts to manage the project.[13] IPTO was combined with the Transformational Convergence Technology Office (TCTO) to form the Information Innovation Office (I2O) in 2010." two immediate things came to mind: #1.) it would seem of benefit that this part of IPTO's history could/needs to be further flushed out to include all prior directors and their assistants, say starting with Bob Kahn, Steve Crocker, ...? #2.) a 26 year old D/ARPA office director (Ivan Sutherland)!? would curious to know if that would that be the youngest (or anything close to the youngest) director of any D/ARPA office to the present day? geoff On Fri, Nov 26, 2021 at 2:59 PM the keyboard of geoff goodfellow < geoff at iconia.com> wrote: > Alex, > > please pardon the always being the unrelenting curious and inquisitive > mind here, but would you be willing to share what were the several specific > items which the publishers considered critical and you were required to > cover in your history of the Internet paper, pretty please? > > geoff > > On Fri, Nov 26, 2021 at 7:28 AM Alex McKenzie via Internet-history < > internet-history at elists.isoc.org> wrote: > >> In 2014 a paper I was hired to write about the history of the Internet >> was published in the 3-volume "Discoveries in Modern Science." I was >> required to cover several specific items which the publishers considered >> critical, whether I thought so or not. Of course, with a restricted word >> limit the big question for an author is what to include and what to leave >> out. I'm sure other authors would have made different choices from mine. >> I was recently granted permission to post a copy of this article on my >> website. If anyone is interested it can be found at http:// >> alexmckenzie.weebly.com/history-of-the-internet.html >> >> Cheers,Alex >> > > -- Geoff.Goodfellow at iconia.com living as The Truth is True From brian.e.carpenter at gmail.com Fri Nov 26 18:07:09 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Sat, 27 Nov 2021 15:07:09 +1300 Subject: [ih] My 2013 paper on the History of the Internet In-Reply-To: <184824475.660661.1637947706565@mail.yahoo.com> References: <184824475.660661.1637947706565.ref@mail.yahoo.com> <184824475.660661.1637947706565@mail.yahoo.com> Message-ID: <6484f5dc-08cf-41f9-4ec5-bff4ea6312b8@gmail.com> "In December 1990 Tim Berners-Lee (1955? ), an American computer scientist working at the European Organization for Nuclear Research (CERN), demonstrated a system..." I've never met anyone more British than Tim, I don't think. By the way, where does the December 1990 date come from? I'm looking at pages 29-31 of Tim's book, and it confirms what I remember - he and Robert Cailliau didn't start showing the Web to other people until the beginning of 1991. I was a beta user of Nicola Pellow's CLI browser (and its killer app, access to the CERN phone book) from early 1991, although I can't put an exact date on it. Regards Brian Carpenter On 27-Nov-21 06:28, Alex McKenzie via Internet-history wrote: > In 2014 a paper I was hired to write about the history of the Internet was published in the 3-volume "Discoveries in Modern Science." I was required to cover several specific items which the publishers considered critical, whether I thought so or not. Of course, with a restricted word limit the big question for an author is what to include and what to leave out.? I'm sure other authors would have made different choices from mine. > I was recently granted permission to post a copy of this article on my website.? If anyone is interested it can be found athttp://alexmckenzie.weebly.com/history-of-the-internet.html > > Cheers,Alex > From jack at 3kitty.org Fri Nov 26 19:48:49 2021 From: jack at 3kitty.org (Jack Haverty) Date: Fri, 26 Nov 2021 19:48:49 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> Message-ID: <2f562446-079e-1e35-95ef-4759aa4e6a55@3kitty.org> I just ran across a contemporary article on the current use of BGP in the operational Internet: https://krebsonsecurity.com/2021/11/the-internet-is-held-together-with-spit-baling-wire/ Fascinating, to me at least, since I lost track of how BGP was being used back in the 80s when we created it as an interim technique. Honest, I didn't see this article before I sent my historical offering below.?? Looks like "baling wire" is still in the Internet operators' toolbox, after 40 years. Jack Haverty On 11/25/21 12:16 PM, Jack Haverty via Internet-history wrote: > On 11/24/21 5:47 PM, Guy Almes via Internet-history wrote: >> First, by 1987, when both the ARPAnet and the proto-NSFnet backbone >> were both operational, networks that connected to both had to decide >> which to use, and that led to interesting routing decisions. Problems >> encountered then led, for example, to creation of BGP. > > FYI, for The Historians, I offer a little of the earlier history, all > from the 1980-1983 timeframe. > > At the ICCB (before it was renamed IAB) meetings, there was a list > kept on the whiteboard of "things that need to get figured out". One > of the items on that list was "Routing", which included issues like > "Type Of Service", "Shortest Path First", and "Policy Routing" -- all > part of the "How should routing behave in the future Internet?" > topic.? There were two concrete motivators for these issues. > > The Internet had started to evolve from its "fuzzy peach" stage, where > essentially the ARPANET was the peach surrounded by a fuzzy layer of? > LANs, into a richer topology where there were actually choices to be > made. > > First, SATNET had linked the US and Europe for a while, but Bob Kahn > initiated the addition of a transatlantic path using the public X.25 > service.?? The interconnect was called a "VAN Gateway" (VAN stood for > Value Added Network but I never did understand what that really > meant). ? The VAN gateway essentially added an interface option to the > existing gateways, allowing them to connect to the public X.25 > networks, and use them as a "circuit" (albeit virtual) between > gateways.? In effect, the entire X.25 public network system was made > an available component of the Internet; it became just another network > that could be used within the overall Internet. > > The "dial up" nature of the X.25 service also introduced the > possibility of dynamic configuration of the Internet topology -- e.g., > adding or deleting circuits between pairs of gateways as a situation > warranted, simply by using the dialup infrastructure of the public > X.27/X.75 system.? We called that something like "Dynamic Adaptive > Topology", but I don't recall ever actually trying to use that > capability except on the single US<->EU path in parallel with SATNET. > > This "VAN" capability was used to create a topology where there were > two ways IP datagrams could cross the Atlantic.?? Bringing economics > into the picture, the SATNET path was funded by ARPA, to be used only > for ARPA-approved projects.?? The X.25 path was funded by whoever > opened the circuit first (which we, as ARPA contractors, silently > engineered to be usually the European side; seemed like the right > thing to do).?? This issue appeared on the ICCB's to-do list as > "Policy Based Routing". > > Pending the "real" solution, I don't recall exactly how the gateways > back then made the choice of path for each packet; my vague > recollection is that it had something to do with destination addresses > - i.e., a host might have two distinct IP addresses, one for use via > SATNET and the other for use via X.25.?? And a single physical net > would be assigned two network numbers (no shortage then), one for use > via X.25 and the other for use via SATNET. IIRC, that's how the UCL > network in London was configured.? The early Internet had to sometimes > use patching plaster and baling wire to keep it going as the research > sought the right way for the future. > > Second, the Wideband Net, a satellite-based network spanning the US, > was made part of the Internet topology by gateways between it and the > ARPANET.?? There were then multiple network paths across the US.?? But > the gateways' routing metric of "hops" would never cause any datagrams > to be sent over the Wideband Net.?? Since the Wideband Net was only > interconnected by gateways to ARPANET nodes, any route that used the > Wideband Net would necessarily be 2 hops longer than a direct path > across the ARPANET.?? The routing mechanisms would never make such a > choice.??? This issue was captured on the ICCB's list as "Expressway > Routing", a reference to the need for car drivers to make a decision > to head toward the nearest freeway entrance, rather than taking a > direct route toward their destination, in order to get to their > destination faster. > > I don't recall how people experimented with the Wideband Net, i.e., > how they got datagrams to actually flow over it.?? Perhaps that was a > use of the "Source Routing" mechanisms in the IP headers.?? Maybe > someone else remembers.... > > We didn't know how to best address these situations, but of course > there were a lot of ideas.? In addition, the existing Internet lacked > some basic mechanisms that seemed to be necessary.? In particular, the > use of "hops" to determine which path was the shortest was woefully > inadequate.?? A "hop" through a Satellite net might be expected to > take much longer than a hop through a terrestrial net, simply due to > Physics. ? But a hop through the ARPANET traversing many IMPs, when > the net was congested, might actually take longer than a satellite > transit.? A time-based metric was not feasible in the gateways without > some means of accurately measuring time, at a precision of > milliseconds, in the routers scattered across the continents. > > Dave Mills was on the ICCB, and he took on this quest with unbridled > energy and determination.?? NTP was the result - an impressive piece > of engineering.?? Using NTP, computers (e.g., routers, gateways, > hosts, servers, whatever you call them) can maintain highly > synchronized clocks, and measure actual transit times of IP datagrams > for use in calculating "shortest path". Everyone can thank Dave and > his crew that your computers know what time it is today. > > The Time mechanisms would be helpful, but much more was needed to > handle "Policy Based" and "Expressway" routing situations.? Lots of > people had ideas and wanted them put into the "core gateways" that BBN > operated.?? But doing that kind of experimentation and also keeping > the Internet core reliably running 24x7 was a struggle. > > I was also on the ICCB at the time, and I recruited Dr. Eric Rosen > back at BBN to help think about it.?? He and I had numerous day-long > sessions with a whiteboard.? The result was EGP - the Exterior Gateway > Protocol.?? If you read Eric's now ancient RFC defining EGP, you'll > see that it was not intended as a routing protocol.?? Rather it was > more of a "firewall" mechanism that would allow the Internet to be > carved up into pieces, each of which was implemented and operated at > arm's length from the others but could interoperate to present a > single Internet to the end users. > > The intent was that such a mechanism would make it possible for some > collection of gateways (e.g., the "core gateways" of the Internet at > that time) to be operated as a reliable service, while also enabling > lots of other collections of gateways to be used as guinea pigs for > all sorts of experiments to try out various ideas that had come up.?? > Each such collection was called an "Autonomous System" of gateways > using some particular technical mechanisms and under some single > operator's control.?? EGP was a mechanism to permit reliable > operational services to coexist in the Internet with research and > experimentation. > > When the ideas had been tried, and the traditional "rough consensus" > emerged to identify the best system design, the new algorithms, > mechanisms, protocols, and anything else needed, would be instantiated > in a new Autonomous System, which would then grow as the new system > was deployed - much as the ARPANET has served as the nursery for the > fledgling Internet, with all IMPs disappearing over time as they were > replaced by routers directly connected with wires. > > That's where my direct involvement in the "research" stopped, as I > went more into deploying and operating network stuff, from about > mid-1983 on.?? Perhaps someone else can fill in more gaps in the History. > > Enjoy, > Jack Haverty > > > > From brian.e.carpenter at gmail.com Fri Nov 26 20:11:27 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Sat, 27 Nov 2021 17:11:27 +1300 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <2f562446-079e-1e35-95ef-4759aa4e6a55@3kitty.org> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> <2f562446-079e-1e35-95ef-4759aa4e6a55@3kitty.org> Message-ID: <6c755374-39ed-a360-ad35-0d979c71eb1e@gmail.com> If you want to be even more concerned, see the article about securing BGP at page 19 of: https://ipj.dreamhosters.com/wp-content/uploads/2021/10/243-ipj.pdf Regards Brian On 27-Nov-21 16:48, Jack Haverty via Internet-history wrote: > I just ran across a contemporary article on the current use of BGP in > the operational Internet: > > https://krebsonsecurity.com/2021/11/the-internet-is-held-together-with-spit-baling-wire/ > > Fascinating, to me at least, since I lost track of how BGP was being > used back in the 80s when we created it as an interim technique. Honest, > I didn't see this article before I sent my historical offering below. > Looks like "baling wire" is still in the Internet operators' toolbox, > after 40 years. > > Jack Haverty > > > On 11/25/21 12:16 PM, Jack Haverty via Internet-history wrote: >> On 11/24/21 5:47 PM, Guy Almes via Internet-history wrote: >>> First, by 1987, when both the ARPAnet and the proto-NSFnet backbone >>> were both operational, networks that connected to both had to decide >>> which to use, and that led to interesting routing decisions. Problems >>> encountered then led, for example, to creation of BGP. >> >> FYI, for The Historians, I offer a little of the earlier history, all >> from the 1980-1983 timeframe. >> >> At the ICCB (before it was renamed IAB) meetings, there was a list >> kept on the whiteboard of "things that need to get figured out". One >> of the items on that list was "Routing", which included issues like >> "Type Of Service", "Shortest Path First", and "Policy Routing" -- all >> part of the "How should routing behave in the future Internet?" >> topic.? There were two concrete motivators for these issues. >> >> The Internet had started to evolve from its "fuzzy peach" stage, where >> essentially the ARPANET was the peach surrounded by a fuzzy layer of >> LANs, into a richer topology where there were actually choices to be >> made. >> >> First, SATNET had linked the US and Europe for a while, but Bob Kahn >> initiated the addition of a transatlantic path using the public X.25 >> service.?? The interconnect was called a "VAN Gateway" (VAN stood for >> Value Added Network but I never did understand what that really >> meant). ? The VAN gateway essentially added an interface option to the >> existing gateways, allowing them to connect to the public X.25 >> networks, and use them as a "circuit" (albeit virtual) between >> gateways.? In effect, the entire X.25 public network system was made >> an available component of the Internet; it became just another network >> that could be used within the overall Internet. >> >> The "dial up" nature of the X.25 service also introduced the >> possibility of dynamic configuration of the Internet topology -- e.g., >> adding or deleting circuits between pairs of gateways as a situation >> warranted, simply by using the dialup infrastructure of the public >> X.27/X.75 system.? We called that something like "Dynamic Adaptive >> Topology", but I don't recall ever actually trying to use that >> capability except on the single US<->EU path in parallel with SATNET. >> >> This "VAN" capability was used to create a topology where there were >> two ways IP datagrams could cross the Atlantic.?? Bringing economics >> into the picture, the SATNET path was funded by ARPA, to be used only >> for ARPA-approved projects.?? The X.25 path was funded by whoever >> opened the circuit first (which we, as ARPA contractors, silently >> engineered to be usually the European side; seemed like the right >> thing to do).?? This issue appeared on the ICCB's to-do list as >> "Policy Based Routing". >> >> Pending the "real" solution, I don't recall exactly how the gateways >> back then made the choice of path for each packet; my vague >> recollection is that it had something to do with destination addresses >> - i.e., a host might have two distinct IP addresses, one for use via >> SATNET and the other for use via X.25.?? And a single physical net >> would be assigned two network numbers (no shortage then), one for use >> via X.25 and the other for use via SATNET. IIRC, that's how the UCL >> network in London was configured.? The early Internet had to sometimes >> use patching plaster and baling wire to keep it going as the research >> sought the right way for the future. >> >> Second, the Wideband Net, a satellite-based network spanning the US, >> was made part of the Internet topology by gateways between it and the >> ARPANET.?? There were then multiple network paths across the US.?? But >> the gateways' routing metric of "hops" would never cause any datagrams >> to be sent over the Wideband Net.?? Since the Wideband Net was only >> interconnected by gateways to ARPANET nodes, any route that used the >> Wideband Net would necessarily be 2 hops longer than a direct path >> across the ARPANET.?? The routing mechanisms would never make such a >> choice.??? This issue was captured on the ICCB's list as "Expressway >> Routing", a reference to the need for car drivers to make a decision >> to head toward the nearest freeway entrance, rather than taking a >> direct route toward their destination, in order to get to their >> destination faster. >> >> I don't recall how people experimented with the Wideband Net, i.e., >> how they got datagrams to actually flow over it.?? Perhaps that was a >> use of the "Source Routing" mechanisms in the IP headers.?? Maybe >> someone else remembers.... >> >> We didn't know how to best address these situations, but of course >> there were a lot of ideas.? In addition, the existing Internet lacked >> some basic mechanisms that seemed to be necessary.? In particular, the >> use of "hops" to determine which path was the shortest was woefully >> inadequate.?? A "hop" through a Satellite net might be expected to >> take much longer than a hop through a terrestrial net, simply due to >> Physics. ? But a hop through the ARPANET traversing many IMPs, when >> the net was congested, might actually take longer than a satellite >> transit.? A time-based metric was not feasible in the gateways without >> some means of accurately measuring time, at a precision of >> milliseconds, in the routers scattered across the continents. >> >> Dave Mills was on the ICCB, and he took on this quest with unbridled >> energy and determination.?? NTP was the result - an impressive piece >> of engineering.?? Using NTP, computers (e.g., routers, gateways, >> hosts, servers, whatever you call them) can maintain highly >> synchronized clocks, and measure actual transit times of IP datagrams >> for use in calculating "shortest path". Everyone can thank Dave and >> his crew that your computers know what time it is today. >> >> The Time mechanisms would be helpful, but much more was needed to >> handle "Policy Based" and "Expressway" routing situations.? Lots of >> people had ideas and wanted them put into the "core gateways" that BBN >> operated.?? But doing that kind of experimentation and also keeping >> the Internet core reliably running 24x7 was a struggle. >> >> I was also on the ICCB at the time, and I recruited Dr. Eric Rosen >> back at BBN to help think about it.?? He and I had numerous day-long >> sessions with a whiteboard.? The result was EGP - the Exterior Gateway >> Protocol.?? If you read Eric's now ancient RFC defining EGP, you'll >> see that it was not intended as a routing protocol.?? Rather it was >> more of a "firewall" mechanism that would allow the Internet to be >> carved up into pieces, each of which was implemented and operated at >> arm's length from the others but could interoperate to present a >> single Internet to the end users. >> >> The intent was that such a mechanism would make it possible for some >> collection of gateways (e.g., the "core gateways" of the Internet at >> that time) to be operated as a reliable service, while also enabling >> lots of other collections of gateways to be used as guinea pigs for >> all sorts of experiments to try out various ideas that had come up. >> Each such collection was called an "Autonomous System" of gateways >> using some particular technical mechanisms and under some single >> operator's control.?? EGP was a mechanism to permit reliable >> operational services to coexist in the Internet with research and >> experimentation. >> >> When the ideas had been tried, and the traditional "rough consensus" >> emerged to identify the best system design, the new algorithms, >> mechanisms, protocols, and anything else needed, would be instantiated >> in a new Autonomous System, which would then grow as the new system >> was deployed - much as the ARPANET has served as the nursery for the >> fledgling Internet, with all IMPs disappearing over time as they were >> replaced by routers directly connected with wires. >> >> That's where my direct involvement in the "research" stopped, as I >> went more into deploying and operating network stuff, from about >> mid-1983 on.?? Perhaps someone else can fill in more gaps in the History. >> >> Enjoy, >> Jack Haverty >> >> >> >> > > From johnl at iecc.com Fri Nov 26 20:46:13 2021 From: johnl at iecc.com (John Levine) Date: 26 Nov 2021 23:46:13 -0500 Subject: [ih] My 2013 paper on the History of the Internet In-Reply-To: <6484f5dc-08cf-41f9-4ec5-bff4ea6312b8@gmail.com> Message-ID: <20211127044614.7B705309B73D@ary.qy> It appears that Brian E Carpenter via Internet-history said: >"In December 1990 Tim Berners-Lee (1955? ), an >American computer scientist working at the European >Organization for Nuclear Research (CERN), demonstrated >a system..." > >I've never met anyone more British than Tim, I don't think. Probably sloppy reporting. Tim has lived in the US for a long time so he must be American. R's, John From tony.li at tony.li Fri Nov 26 23:14:22 2021 From: tony.li at tony.li (Tony Li) Date: Fri, 26 Nov 2021 23:14:22 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <6c755374-39ed-a360-ad35-0d979c71eb1e@gmail.com> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> <2f562446-079e-1e35-95ef-4759aa4e6a55@3kitty.org> <6c755374-39ed-a360-ad35-0d979c71eb1e@gmail.com> Message-ID: <61D05837-0508-4154-B5D1-75D923865E99@tony.li> Then take a look at https://rpki-monitor.antd.nist.gov/ and see that we?re actually making pretty steady progress on getting RPKI deployed. Tony > On Nov 26, 2021, at 8:11 PM, Brian E Carpenter via Internet-history wrote: > > If you want to be even more concerned, see the article about securing BGP > at page 19 of: > https://ipj.dreamhosters.com/wp-content/uploads/2021/10/243-ipj.pdf > > Regards > Brian > > On 27-Nov-21 16:48, Jack Haverty via Internet-history wrote: >> I just ran across a contemporary article on the current use of BGP in >> the operational Internet: >> https://krebsonsecurity.com/2021/11/the-internet-is-held-together-with-spit-baling-wire/ >> Fascinating, to me at least, since I lost track of how BGP was being >> used back in the 80s when we created it as an interim technique. Honest, >> I didn't see this article before I sent my historical offering below. >> Looks like "baling wire" is still in the Internet operators' toolbox, >> after 40 years. >> Jack Haverty >> On 11/25/21 12:16 PM, Jack Haverty via Internet-history wrote: >>> On 11/24/21 5:47 PM, Guy Almes via Internet-history wrote: >>>> First, by 1987, when both the ARPAnet and the proto-NSFnet backbone >>>> were both operational, networks that connected to both had to decide >>>> which to use, and that led to interesting routing decisions. Problems >>>> encountered then led, for example, to creation of BGP. >>> >>> FYI, for The Historians, I offer a little of the earlier history, all >>> from the 1980-1983 timeframe. >>> >>> At the ICCB (before it was renamed IAB) meetings, there was a list >>> kept on the whiteboard of "things that need to get figured out". One >>> of the items on that list was "Routing", which included issues like >>> "Type Of Service", "Shortest Path First", and "Policy Routing" -- all >>> part of the "How should routing behave in the future Internet?" >>> topic. There were two concrete motivators for these issues. >>> >>> The Internet had started to evolve from its "fuzzy peach" stage, where >>> essentially the ARPANET was the peach surrounded by a fuzzy layer of >>> LANs, into a richer topology where there were actually choices to be >>> made. >>> >>> First, SATNET had linked the US and Europe for a while, but Bob Kahn >>> initiated the addition of a transatlantic path using the public X.25 >>> service. The interconnect was called a "VAN Gateway" (VAN stood for >>> Value Added Network but I never did understand what that really >>> meant). The VAN gateway essentially added an interface option to the >>> existing gateways, allowing them to connect to the public X.25 >>> networks, and use them as a "circuit" (albeit virtual) between >>> gateways. In effect, the entire X.25 public network system was made >>> an available component of the Internet; it became just another network >>> that could be used within the overall Internet. >>> >>> The "dial up" nature of the X.25 service also introduced the >>> possibility of dynamic configuration of the Internet topology -- e.g., >>> adding or deleting circuits between pairs of gateways as a situation >>> warranted, simply by using the dialup infrastructure of the public >>> X.27/X.75 system. We called that something like "Dynamic Adaptive >>> Topology", but I don't recall ever actually trying to use that >>> capability except on the single US<->EU path in parallel with SATNET. >>> >>> This "VAN" capability was used to create a topology where there were >>> two ways IP datagrams could cross the Atlantic. Bringing economics >>> into the picture, the SATNET path was funded by ARPA, to be used only >>> for ARPA-approved projects. The X.25 path was funded by whoever >>> opened the circuit first (which we, as ARPA contractors, silently >>> engineered to be usually the European side; seemed like the right >>> thing to do). This issue appeared on the ICCB's to-do list > as >>> "Policy Based Routing". >>> >>> Pending the "real" solution, I don't recall exactly how the gateways >>> back then made the choice of path for each packet; my vague >>> recollection is that it had something to do with destination addresses >>> - i.e., a host might have two distinct IP addresses, one for use via >>> SATNET and the other for use via X.25. And a single physical net >>> would be assigned two network numbers (no shortage then), one for use >>> via X.25 and the other for use via SATNET. IIRC, that's how the UCL >>> network in London was configured. The early Internet had to sometimes >>> use patching plaster and baling wire to keep it going as the research >>> sought the right way for the future. >>> >>> Second, the Wideband Net, a satellite-based network spanning the US, >>> was made part of the Internet topology by gateways between it and the >>> ARPANET. There were then multiple network paths across the > US. But >>> the gateways' routing metric of "hops" would never cause any datagrams >>> to be sent over the Wideband Net. Since the Wideband Net was only >>> interconnected by gateways to ARPANET nodes, any route that used the >>> Wideband Net would necessarily be 2 hops longer than a direct path >>> across the ARPANET. The routing mechanisms would never make such a >>> choice. This issue was captured on the ICCB's list as "Expressway >>> Routing", a reference to the need for car drivers to make a decision >>> to head toward the nearest freeway entrance, rather than taking a >>> direct route toward their destination, in order to get to their >>> destination faster. >>> >>> I don't recall how people experimented with the Wideband Net, i.e., >>> how they got datagrams to actually flow over it. Perhaps that was a >>> use of the "Source Routing" mechanisms in the IP headers. Maybe >>> someone else remembers.... >>> >>> We didn't know how to best address these situations, but of course >>> there were a lot of ideas. In addition, the existing Internet lacked >>> some basic mechanisms that seemed to be necessary. In particular, the >>> use of "hops" to determine which path was the shortest was woefully >>> inadequate. A "hop" through a Satellite net might be expected to >>> take much longer than a hop through a terrestrial net, simply due to >>> Physics. But a hop through the ARPANET traversing many IMPs, when >>> the net was congested, might actually take longer than a satellite >>> transit. A time-based metric was not feasible in the gateways without >>> some means of accurately measuring time, at a precision of >>> milliseconds, in the routers scattered across the continents. >>> >>> Dave Mills was on the ICCB, and he took on this quest with unbridled >>> energy and determination. NTP was the result - an impressive piece >>> of engineering. Using NTP, computers (e.g., routers, gateways, >>> hosts, servers, whatever you call them) can maintain highly >>> synchronized clocks, and measure actual transit times of IP datagrams >>> for use in calculating "shortest path". Everyone can thank Dave and >>> his crew that your computers know what time it is today. >>> >>> The Time mechanisms would be helpful, but much more was needed to >>> handle "Policy Based" and "Expressway" routing situations. Lots of >>> people had ideas and wanted them put into the "core gateways" that BBN >>> operated. But doing that kind of experimentation and also keeping >>> the Internet core reliably running 24x7 was a struggle. >>> >>> I was also on the ICCB at the time, and I recruited Dr. Eric Rosen >>> back at BBN to help think about it. He and I had numerous day-long >>> sessions with a whiteboard. The result was EGP - the Exterior Gateway >>> Protocol. If you read Eric's now ancient RFC defining EGP, > you'll >>> see that it was not intended as a routing protocol. Rather > it was >>> more of a "firewall" mechanism that would allow the Internet to be >>> carved up into pieces, each of which was implemented and operated at >>> arm's length from the others but could interoperate to present a >>> single Internet to the end users. >>> >>> The intent was that such a mechanism would make it possible for some >>> collection of gateways (e.g., the "core gateways" of the Internet at >>> that time) to be operated as a reliable service, while also enabling >>> lots of other collections of gateways to be used as guinea pigs for >>> all sorts of experiments to try out various ideas that had come up. >>> Each such collection was called an "Autonomous System" of gateways >>> using some particular technical mechanisms and under some single >>> operator's control. EGP was a mechanism to permit reliable >>> operational services to coexist in the Internet with research and >>> experimentation. >>> >>> When the ideas had been tried, and the traditional "rough consensus" >>> emerged to identify the best system design, the new algorithms, >>> mechanisms, protocols, and anything else needed, would be instantiated >>> in a new Autonomous System, which would then grow as the new system >>> was deployed - much as the ARPANET has served as the nursery for the >>> fledgling Internet, with all IMPs disappearing over time as they were >>> replaced by routers directly connected with wires. >>> >>> That's where my direct involvement in the "research" stopped, as I >>> went more into deploying and operating network stuff, from about >>> mid-1983 on. Perhaps someone else can fill in more gaps in > the History. >>> >>> Enjoy, >>> Jack Haverty >>> >>> >>> >>> > > -- > Internet-history mailing list > Internet-history at elists.isoc.org > https://elists.isoc.org/mailman/listinfo/internet-history From steve at shinkuro.com Fri Nov 26 23:28:04 2021 From: steve at shinkuro.com (Steve Crocker) Date: Sat, 27 Nov 2021 02:28:04 -0500 Subject: [ih] ARPA/IPTO personnel history In-Reply-To: References: <184824475.660661.1637947706565.ref@mail.yahoo.com> <184824475.660661.1637947706565@mail.yahoo.com> Message-ID: On Fri, Nov 26, 2021 at 8:24 PM the keyboard of geoff goodfellow via Internet-history wrote: > in reading Alex's 2013 paper on the History of the Internet yours truly saw mention > as to who was the first IPTO director (Lick) and then immediate mention > that Robert W. ??Bob?? Taylor (1932? ) became the third IPTO director in > 1966 -- but with no mention as to who was the second IPTO director... so > over to Wikipedia it was and yours truly found at > > https://en.wikipedia.org/wiki/Information_Processing_Techniques_Office > > under "Later history > > Ivan Sutherland replaced J. C. R. Licklider as IPTO's director when Licklider > left ARPA in 1964.[7][8] Sutherland was 26 years old at the time. > Bob Taylor was hired as Sutherland's assistant in 1965 and became director in > 1966.[9] > > #1.) it would seem of benefit that this part of IPTO's history could/needs to > be further flushed out to include all prior directors and their > assistants, say starting with Bob Kahn, Steve Crocker, ...? > Even though IPTO was a small office, it would take a fair bit of work to assemble the IPTO personnel history. I was in the office for only three years, mid 1971 to mid 1974. Here's as much of the personnel history as I can recall quickly from memory. Directors: 1. JCR Licklider 2. Ivan Sutherland 3. Bob Taylor 4. Larry Roberts 5. Lick again 6. Dave Russell 7. Bob Kahn ... In some order that I don't have handy, - Saul Amaral - Jack Schwartz - ,,, Almost all of the non-clerical people in the office were program managers. A couple of exceptions are noted below. Program Managers Art Bushkin Barry Wessler Cordell Green Bruce Dolan John Perry Steve Crocker Bob Kahn Peggy Karp Dave Carlstrom Bill Carlson Steve Walker Craig Fields Vint Cerf Duane Adams Ron Ohlander Steve Squires Larry Druffel Bill Scherlis Brian Boesch Hilarie Orman Kirstie Bellman and many more. While I was in the office, Al Blue was a senior non-technical guy who handled our budget, the interactions with the contracting process and generally kept things running smoothly. He was super smart. self effacing and a delight to work with. We often said if all of us technical people disappeared, he could keep the place running for quite a while based on his knowledge of the budget, the track record of the principal investigators, and his skill in dealing with the government bureaucracy. He was a history buff and I believe he wrote a book about B-24s in World War II. Gene Stubbs was another senior non-technical person we recruited from one of the contracting ships. My apologies to anyone I left out. Steve From dhc at dcrocker.net Sat Nov 27 05:52:52 2021 From: dhc at dcrocker.net (Dave Crocker) Date: Sat, 27 Nov 2021 05:52:52 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <61D05837-0508-4154-B5D1-75D923865E99@tony.li> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> <2f562446-079e-1e35-95ef-4759aa4e6a55@3kitty.org> <6c755374-39ed-a360-ad35-0d979c71eb1e@gmail.com> <61D05837-0508-4154-B5D1-75D923865E99@tony.li> Message-ID: <0db831bd-f52d-ad0e-8eb1-73977d52386f@dcrocker.net> On 11/26/2021 11:14 PM, Tony Li via Internet-history wrote: > Then take a look athttps://rpki-monitor.antd.nist.gov/ and see that we?re actually making pretty steady progress on > getting RPKI deployed. It would be helpful to see a curve of adoption over time, from the first production-ready version of the RPKI spec. The site doesn't offer that (and it only goes back to April, 2019.) d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net From jnc at mercury.lcs.mit.edu Sat Nov 27 06:53:27 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Sat, 27 Nov 2021 09:53:27 -0500 (EST) Subject: [ih] "The First Router" on Jeopardy Message-ID: <20211127145327.EFAF718C07B@mercury.lcs.mit.edu> > From: Scott Brim > I'm just going to toss something in here from 1651. .... There is > nothing new under the sun etc. Indeed; along similar lines is this quotation: "I am far from thinking that nomenclature is a remedy for every defect in art or science: still I cannot but feel that confusion of terms generally springs from, and always leads to, confusion of ideas." -- John Louis Petit, "Architectural Studies in France", 1854 which a friend mine found several decades (sigh) back. To your last observation, this one is 'only' from 1854! Noel From tony.li at tony.li Sat Nov 27 07:11:56 2021 From: tony.li at tony.li (Tony Li) Date: Sat, 27 Nov 2021 07:11:56 -0800 Subject: [ih] "The First Router" on Jeopardy In-Reply-To: <0db831bd-f52d-ad0e-8eb1-73977d52386f@dcrocker.net> References: <315942109.667960.1637767799111@mail.yahoo.com> <83ef0fe4-a2c7-4e8e-0783-3224409c4a0e@tamu.edu> <2f562446-079e-1e35-95ef-4759aa4e6a55@3kitty.org> <6c755374-39ed-a360-ad35-0d979c71eb1e@gmail.com> <61D05837-0508-4154-B5D1-75D923865E99@tony.li> <0db831bd-f52d-ad0e-8eb1-73977d52386f@dcrocker.net> Message-ID: > On Nov 27, 2021, at 5:52 AM, Dave Crocker wrote: > > > It would be helpful to see a curve of adoption over time, from the first production-ready version of the RPKI spec. > > The site doesn't offer that (and it only goes back to April, 2019.) The glass is a third full. "He has a right to criticize who has a heart to help." --Abraham Lincoln Tony From jnc at mercury.lcs.mit.edu Sat Nov 27 08:06:07 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Sat, 27 Nov 2021 11:06:07 -0500 (EST) Subject: [ih] multi-protocol routers, bridges (Was: "The First Router" on Jeopardy) Message-ID: <20211127160607.1798E18C07B@mercury.lcs.mit.edu> > Bill Nowicki > As we soon ... cloned the MIT LSI-11 system This will not really be of interest to people here (but maybe - see the Len Bosack story below), but I wanted to put a corrrection in the record close to the original. I don't think Stanford actually "cloned the MIT LSI-11 system" (probably because doing so would have involved getting the toolchain, which ran on the MIT V6 PWB1 PDP-11 UNIX, running there); instead, loads for the one router at Stanford which ran it - "Golden", Stanford's ARPANET gateway - were built at MIT. I recall making at least one trip out to Stanford to work on Golden (installing metering to monitor packet queue lengths, IIRC), during which I ran into Len Bosack, who was running the Stanford timesharing systems at that point, and explained to him how there was going to be a big maket for routers, and I was in the process of working with Proteon to turn out a product. In retrospect, probably not the best idea! :-) Jeff Mogul may also have been involved; I found his name in some log files (below). He'd been associated with our group at LCS as an undergrad, when he put a ring interface on the DSSR/RTS VAX; he then went to grad school at Stanford. He would have known of the router work at MIT, and probably they figured it was easier to use our code on their ARPANET gateway, than write ARPANET code of their own for Bill Yeager's router. I have a dump of that MIT V6 PWB1 PDP-11 UNIX, and just looked at the config files. Below are some amusing/interesting snippes from a later one, after a DIX Ethernet board had been added to it. Interestingly, although there was PUP code available for the MIT router, Golden seems to have been IP-only (from what I can see in the config files). Noel ---- /* Node name */ char name[] "Golden-Gateway"; /* Actual network configuration table. Note the trailer word in * the ARPANet line; NOT a hardware problem in ANY board, but * IMP trailer. (See your 1822 manual, tuna.) * LOSERS: don't frobozz the n_max packet count fields; this * configuration is intended to maximize ARPANet throughput * in a gateway that's not computationally loaded. In particular, * leave the ARPANet maxip larger than the Ethernet!!! * "We like our defaults. If we hadn't liked them, we wouldn't * have made them the defaults, now, would we." */ { { NULL, ð_prinit, ð_out, NULL, ð_in, NULL, NULL, /* EtherNet */ 2, 3, 4, 0, ETHMAX, sizeof(struct ethpkt), 0, DVETH, DVETH+1, DVETH, T_ETH, C_BRD }, { &ar_init, &ar_prinit, &ar_out, NULL, NULL, &ar_get, NULL, /* ARPANet */ 4, 3, 0, 0, ARMAX, sizeof(arpkt), sizeof(word), DVIMP, DVIMP+1, DVIMP, T_ARPA, 0 }, { NULL, &e10_prinit, &e10_out, NULL, &e10_in, NULL, NULL, /* 10MB Ethernet */ 1, 3, 4, 0, EMAX, sizeof(e10pkt), 0, DVIE, DVIE+1, DVIE, T_E10, C_BRD }, }; #define NINADDR 3 inia iniatbl[NINADDR] { { { 36, 050, 0, 076 }, NULL }, /* Ethernet */ { { 012, 1, 0, 11 }, NULL }, /* ARPAnet */ { { 36, 8, 0, 1 }, NULL }, /* 10MB Ethernet */ }; /* For subnet kludgery; if you are on a class A net with subnets set * this variable to your net number to enable subnet routing. If * zero, this feature is turned off. */ unsb myanet 36; From amckenzie3 at yahoo.com Sat Nov 27 08:48:58 2021 From: amckenzie3 at yahoo.com (Alex McKenzie) Date: Sat, 27 Nov 2021 16:48:58 +0000 (UTC) Subject: [ih] My 2013 paper on the History of the Internet In-Reply-To: <6484f5dc-08cf-41f9-4ec5-bff4ea6312b8@gmail.com> References: <184824475.660661.1637947706565.ref@mail.yahoo.com> <184824475.660661.1637947706565@mail.yahoo.com> <6484f5dc-08cf-41f9-4ec5-bff4ea6312b8@gmail.com> Message-ID: <1046461577.1391697.1638031738315@mail.yahoo.com> After I submitted my article the editors decided to put all references to individuals in a standard format.? They made Tim an American - I didn't get a chance to review these editorial changes.? I don't remember where the December 1990 reference came from, but I was confident of it at the time I wrote the article. Cheers,Alex On Friday, November 26, 2021, 09:07:15 PM EST, Brian E Carpenter wrote: "In December 1990 Tim Berners-Lee (1955? ), an American computer scientist working at the European Organization for Nuclear Research (CERN), demonstrated a system..." I've never met anyone more British than Tim, I don't think. By the way, where does the December 1990 date come from? I'm looking at pages 29-31 of Tim's book, and it confirms what I remember - he and Robert Cailliau didn't start showing the Web to other people until the beginning of 1991. I was a beta user of Nicola Pellow's CLI browser (and its killer app, access to the CERN phone book) from early 1991, although I can't put an exact date on it. Regards ? ? Brian Carpenter On 27-Nov-21 06:28, Alex McKenzie via Internet-history wrote: > In 2014 a paper I was hired to write about the history of the Internet was published in the 3-volume "Discoveries in Modern Science." I was required to cover several specific items which the publishers considered critical, whether I thought so or not. Of course, with a restricted word limit the big question for an author is what to include and what to leave out.? I'm sure other authors would have made different choices from mine. > I was recently granted permission to post a copy of this article on my website.? If anyone is interested it can be found athttp://alexmckenzie.weebly.com/history-of-the-internet.html > > Cheers,Alex > From winowicki at yahoo.com Sat Nov 27 10:02:28 2021 From: winowicki at yahoo.com (Bill Nowicki) Date: Sat, 27 Nov 2021 18:02:28 +0000 (UTC) Subject: [ih] multi-protocol routers, bridges (Was: "The First Router" on Jeopardy) In-Reply-To: <20211127160607.1798E18C07B@mercury.lcs.mit.edu> References: <20211127160607.1798E18C07B@mercury.lcs.mit.edu> Message-ID: <625879769.2284692.1638036148404@mail.yahoo.com> Yes Noel. Indeed, I meant that we cloned the hardware, and just ran the MIT PDP-11 code to connect to an IMP. which was IP only (although two kinds of Ethernet). The PUP routing was running on what became the Sun hardware (Motorola 68000) with very different software using the PUP routing protocol on the interior of the campus. That is what evolved into the Cisco hardware when Len Bosack and Sandy Lerner formed the company a couple years later. Bill? On Saturday, November 27, 2021, 08:06:14 AM PST, Noel Chiappa via Internet-history wrote: ? ? > Bill Nowicki ? ? > As we soon ... cloned the MIT LSI-11 system This will not really be of interest to people here (but maybe - see the Len Bosack story below), but I wanted to put a corrrection in the record close to the original. I don't think Stanford actually "cloned the MIT LSI-11 system" (probably because doing so would have involved getting the toolchain, which ran on the MIT V6 PWB1 PDP-11 UNIX, running there); instead, loads for the one router at Stanford which ran it - "Golden", Stanford's ARPANET gateway - were built at MIT. I recall making at least one trip out to Stanford to work on Golden (installing metering to monitor packet queue lengths, IIRC), during which I ran into Len Bosack, who was running the Stanford timesharing systems at that point, and explained to him how there was going to be a big maket for routers, and I was in the process of working with Proteon to turn out a product. In retrospect, probably not the best idea! :-) Jeff Mogul may also have been involved; I found his name in some log files (below). He'd been associated with our group at LCS as an undergrad, when he put a ring interface on the DSSR/RTS VAX; he then went to grad school at Stanford. He would have known of the router work at MIT, and probably they figured it was easier to use our code on their ARPANET gateway, than write ARPANET code of their own for Bill Yeager's router. I have a dump of that MIT V6 PWB1 PDP-11 UNIX, and just looked at the config files. Below are some amusing/interesting snippes from a later one, after a DIX Ethernet board had been added to it. Interestingly, although there was PUP code available for the MIT router, Golden seems to have been IP-only (from what I can see in the config files). ??? Noel ---- ? /* Node name */ ? char? ? name[]? "Golden-Gateway"; ? /* Actual network configuration table. Note the trailer word in ? * the ARPANet line; NOT a hardware problem in ANY board, but ? * IMP trailer. (See your 1822 manual, tuna.) ? * LOSERS: don't frobozz the n_max packet count fields; this ? * configuration is intended to maximize ARPANet throughput ? * in a gateway that's not computationally loaded. In particular, ? * leave the ARPANet maxip larger than the Ethernet!!! ? * "We like our defaults. If we hadn't liked them, we wouldn't ? * have made them the defaults, now, would we." ? */ ? { ? { NULL, ð_prinit, ð_out, NULL, ð_in, NULL, NULL, ? ? /* EtherNet */? ? ? 2, 3, 4, 0, ETHMAX, sizeof(struct ethpkt), 0, ??? ??? ??? ? DVETH, DVETH+1, DVETH, T_ETH, C_BRD }, ? { &ar_init, &ar_prinit, &ar_out, NULL, NULL, &ar_get, NULL, ? ? /* ARPANet */? ? ? ? 4, 3, 0, 0, ARMAX, sizeof(arpkt), sizeof(word), ??? ??? ??? ? DVIMP, DVIMP+1, DVIMP, T_ARPA, 0 }, ? { NULL, &e10_prinit, &e10_out, NULL, &e10_in, NULL, NULL, ? ? /* 10MB Ethernet */? 1, 3, 4, 0, EMAX, sizeof(e10pkt), 0, ??? ??? ??? ? DVIE, DVIE+1, DVIE, T_E10, C_BRD }, ? }; ? #define NINADDR 3 ? inia? ? iniatbl[NINADDR] ? {? ? ? { { 36, 050, 0, 076 }, NULL },? ? ? ? ? /* Ethernet */ ??? ? { { 012, 1, 0, 11 }, NULL },? ? ? ? ? ? /* ARPAnet */ ??? ? { { 36, 8, 0, 1 }, NULL },? ? ? ? ? ? ? /* 10MB Ethernet */ ? }; ? /* For subnet kludgery; if you are on a class A net with subnets set ? * this variable to your net number to enable subnet routing. If ? * zero, this feature is turned off. ? */ ? unsb? ? myanet? 36; -- Internet-history mailing list Internet-history at elists.isoc.org https://elists.isoc.org/mailman/listinfo/internet-history From gregskinner0 at icloud.com Sat Nov 27 14:45:59 2021 From: gregskinner0 at icloud.com (Greg Skinner) Date: Sat, 27 Nov 2021 14:45:59 -0800 Subject: [ih] The term "router" in RFCs (Re: "The First Router" on Jeopardy) In-Reply-To: <3A07A8E7-7696-4CD4-A152-93F322CD755E@tzi.org> References: <3d619a8c-7bc6-3284-e746-66aa759ab5a8@3kitty.org> <3A07A8E7-7696-4CD4-A152-93F322CD755E@tzi.org> Message-ID: If I have more time, I?ll provide more details, but here are a couple of other uses of ?router? from the 1980s I was able to dig up: There was a program called egpup that ran as a daemon on 4.2 BSD UNIX that implemented the EGP protocol. It is described in RFC911 , although egpup doesn?t appear in the text. After googling for more information, the query q=mike+karels+egpup yielded a link to ISI/RR-84-145 an ISI technical report published in October 1984. The first five sections of the report look like a duplicate of RFC911 (based on quickly eyeballing it). The remaining sections are missing from RFC911. They contain logs where ?router? appears frequently. In a previous message, I mentioned BARRnet, the SF Bay Area NSFnet regional. In the June 1987 Internet Monthly Report the BARRnet report makes liberal use of ?router?. The first use of ?router? I found in this series is in the February 1987 report . ?gregbo > On Nov 24, 2021, at 2:54 AM, Carsten Bormann wrote: > > On 2021-11-22, at 21:50, Jack Haverty via Internet-history wrote: >> >> A historical tidbit -- Back in 1977 Ginny's system was called a "gateway", but later was renamed a "router". It's possible that I did the renaming. At BBN we were selling lots of packet switches, and sometimes customers asked for ideas on how to use their LANs in the network. Our sales people would tell them about the research activities, and the role of gateways, TCP/IP , and the Internet. But in many customers' minds that term "gateway" immediately set off alarm bells, because they had prior bad experience with "gateways" in their IBM networks, and didn't want anything to do with more "gateways". So I suggested calling them "routers" instead of "gateways", and suddenly the marketplace became much more willing to listen. > > A quick look at using the word ?router? in RFCs/IENs: > > RFC 753 (= IEN 85, March 1979) used the word for a message router, as did its update RFC 759 (= IEN 113, August 1980). RFC 850 (Usenet messages, June 1983) used ?mail router?. IEN 178, April 1981, uses ?router? once to describe an entity that did network route processing; the somewhat casual usage doesn?t necessarily imply a forwarding function being part of the ?router? concept. > > The first trace I can find of an Internet gateway being called a router in an RFC is RFC 898 (Gateway SIG Meeting Notes, April 1984): This talks about the CMU gateway, including the lines: > > History - > o "Logical-Host" multiplexor (March 81) > o Gateway (Oct 82) remote debugger and monitor > o Router (Oct 83) > - Modular device and protocol support > - Stub IP dynamic routing > - Local inter-network cable routing. > > Apparently, ?router? implied more functionality than a simple ?gateway?, and ?Stub IP dynamic routing? and ?local inter-network cable routing" seemed to be parts of that. > > So this quick look seems to locate early use of the term with today?s meaning in the early 1980s. > > Gr??e, Carsten > > From jnc at mercury.lcs.mit.edu Sat Nov 27 15:32:00 2021 From: jnc at mercury.lcs.mit.edu (Noel Chiappa) Date: Sat, 27 Nov 2021 18:32:00 -0500 (EST) Subject: [ih] More terminology (Was: multi-protocol routers, bridges) Message-ID: <20211127233200.907B018C07B@mercury.lcs.mit.edu> > From: Bill Nowicki > The PUP routing was running on what became the Sun hardware This brings up another terminological point. To the best of my memory, the below wasn't just _my_ style, but was widely used by people discussing the internetworking layer in the IETF, after the Internet started to really take off. Perhaps some others could weigh in, to confirm/comment? (And for all I know, it's still in use; I'm no longer active in networking, so I don't know.) We distinguished between two very different activities which 'routers' performed; the handling of user traffic, which we called 'forwarding', and the computation of routing data/tables (by routing protocols/algorithms), which was often (but not always, IIRC) called 'routing'. (Slightly confusing, I know! :-) Anyway, by "routing" above, I assume you mean what I have denominated as 'forwarding' above. Of course, they also did PUP 'routing' (as above); and some/all of them eventually did IP 'forwarding' too (to provide the IP traffic for which Stanford needed an IP ARPANET router). Noel From brian.e.carpenter at gmail.com Sat Nov 27 16:59:38 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Sun, 28 Nov 2021 13:59:38 +1300 Subject: [ih] More terminology (Was: multi-protocol routers, bridges) In-Reply-To: <20211127233200.907B018C07B@mercury.lcs.mit.edu> References: <20211127233200.907B018C07B@mercury.lcs.mit.edu> Message-ID: Hi Noel, I'd say that this distinction is usually, but not always, made clear in the IETF today, but the RIB and FIB are very distinct concepts. https://datatracker.ietf.org/doc/draft-king-irtf-challenges-in-routing/ may be of interest re some current thinking. Regards Brian On 28-Nov-21 12:32, Noel Chiappa via Internet-history wrote: > > From: Bill Nowicki > > > The PUP routing was running on what became the Sun hardware > > This brings up another terminological point. To the best of my memory, the > below wasn't just _my_ style, but was widely used by people discussing the > internetworking layer in the IETF, after the Internet started to really take > off. Perhaps some others could weigh in, to confirm/comment? (And for all I > know, it's still in use; I'm no longer active in networking, so I don't know.) > > We distinguished between two very different activities which 'routers' > performed; the handling of user traffic, which we called 'forwarding', and the > computation of routing data/tables (by routing protocols/algorithms), which > was often (but not always, IIRC) called 'routing'. (Slightly confusing, I > know! :-) > > > Anyway, by "routing" above, I assume you mean what I have denominated as > 'forwarding' above. Of course, they also did PUP 'routing' (as above); and > some/all of them eventually did IP 'forwarding' too (to provide the IP traffic > for which Stanford needed an IP ARPANET router). > > Noel > From cabo at tzi.org Sat Nov 27 18:46:32 2021 From: cabo at tzi.org (Carsten Bormann) Date: Sun, 28 Nov 2021 03:46:32 +0100 Subject: [ih] More terminology (Was: multi-protocol routers, bridges) In-Reply-To: <20211127233200.907B018C07B@mercury.lcs.mit.edu> References: <20211127233200.907B018C07B@mercury.lcs.mit.edu> Message-ID: <891318B6-6334-4D22-9B7B-6435F3B2ECDA@tzi.org> On 28. Nov 2021, at 00:32, Noel Chiappa via Internet-history wrote: > > We distinguished between two very different activities which 'routers' > performed; the handling of user traffic, which we called 'forwarding', and the > computation of routing data/tables (by routing protocols/algorithms), which > was often (but not always, IIRC) called 'routing'. (Slightly confusing, I > know! :-) Indeed, but both meanings of ?routing? prevail. I?ll call them routing1 and routing2, where routing1 is defined as the combination of routing2 and fowarding. We?ll use routing1 when describing the overall outcome, as in ?xyz does not route that traffic?, or in ?router?. We?ll use routing2 together with forwarding when it comes to how to implement routing1; RIB and FIB are clear examples of distinct concepts relating to routing2 and forwarding. Routing protocols rarely provide forwarding and therefore are routing2. (A router that uses strict source routing or an SDN setup does not do routing2 at all?) The terms control plane and data plane are another attempt to slice this cake; I must admit I don?t know when those gained popularity. Gr??e, Carsten From brian.e.carpenter at gmail.com Sat Nov 27 19:34:57 2021 From: brian.e.carpenter at gmail.com (Brian E Carpenter) Date: Sun, 28 Nov 2021 16:34:57 +1300 Subject: [ih] More terminology (Was: multi-protocol routers, bridges) In-Reply-To: <891318B6-6334-4D22-9B7B-6435F3B2ECDA@tzi.org> References: <20211127233200.907B018C07B@mercury.lcs.mit.edu> <891318B6-6334-4D22-9B7B-6435F3B2ECDA@tzi.org> Message-ID: <8a7d0824-bb87-7bd7-384d-713b1c373096@gmail.com> On 28-Nov-21 15:46, Carsten Bormann via Internet-history wrote: > On 28. Nov 2021, at 00:32, Noel Chiappa via Internet-history wrote: >> >> We distinguished between two very different activities which 'routers' >> performed; the handling of user traffic, which we called 'forwarding', and the >> computation of routing data/tables (by routing protocols/algorithms), which >> was often (but not always, IIRC) called 'routing'. (Slightly confusing, I >> know! :-) > > Indeed, but both meanings of ?routing? prevail. > I?ll call them routing1 and routing2, where routing1 is defined as the combination of routing2 and fowarding. > > We?ll use routing1 when describing the overall outcome, as in ?xyz does not route that traffic?, or in ?router?. > > We?ll use routing2 together with forwarding when it comes to how to implement routing1; RIB and FIB are clear examples of distinct concepts relating to routing2 and forwarding. Routing protocols rarely provide forwarding and therefore are routing2. > (A router that uses strict source routing or an SDN setup does not do routing2 at all?) > > The terms control plane and data plane are another attempt to slice this cake; I must admit I don?t know when those gained popularity. There's also "fast path" and "slow path" (roughly, hardware vs firmware), a popular distinction when discussing tricky issues like handling IPv6 extension headers at line speed. It's also fairly useless for protocol design, because every router vendor has a different approach to this distinction, as far as I can tell. Brian > > Gr??e, Carsten > From gregskinner0 at icloud.com Mon Nov 29 20:54:05 2021 From: gregskinner0 at icloud.com (Greg Skinner) Date: Mon, 29 Nov 2021 20:54:05 -0800 Subject: [ih] Interop historical material? In-Reply-To: References: <621A290C-8D6E-48DB-B6E2-4726F532D1AE@lynch.com> Message-ID: <9368EFDC-7DAC-4E41-A2D7-240FE57321FA@icloud.com> I have an anecdote from the Interop held in Monterey, CA during March 1987 where routers were mentioned. Dave Clark was one of the plenary speakers. Following the quoted text from the Network World article , ?we?re never going to build networks that work?, I believe he added ?routers that do.? After he finished, Bill Carrico took the stage and said words to the effect that despite problems, it works. DDC then retook the stage and admitted words to the effect that he had been a bit rough. ?gregbo. > On Nov 22, 2021, at 6:26 PM, Jack Haverty wrote: > > Speaking of Interop.... > > Has any of the material from the Interops been preserved??? There were > administrative pieces, like lists of vendors, floor maps, titles of > sessions, speakers' presentations, attendee lists, and such as well as > mountains of presentation slides, papers, vendor handouts, and > promotional material. > > I still have an "Advanced Computing Environments" tile, from before it > was called interop, and one of the "Eyeball" giveaways we handed out > when we did the "Maze" game on the live show net.? ?? Also I even have a > few audiotapes of sessions I gave, although I no longer have any device > that can play them. > > All of this kind of stuff might be historically interesting to someone > and useful as a record of what happened when, especially in the > "commercial" world rather than the "research" world captured in things > like RFCs. > > Jack > > > On 11/22/21 2:37 PM, Dan Lynch via Internet-history wrote: >> And all of them were at Interop 89 in SanJose. Now that?s commercial! >> >> Dan >> >> Cell 650-776-7313 >> >>> On Nov 22, 2021, at 1:36 PM, vinton cerf via Internet-history wrote: >>> >>> ?interestingly 1989 is the arrival of commercial internet (uunet, psinet, >>> cerfnet). >>> >>> v >>> >>> >>>> On Mon, Nov 22, 2021 at 4:14 PM Dave Crocker via Internet-history < >>>> internet-history at elists.isoc.org> wrote: >>>> >>>>> On 11/22/2021 12:50 PM, Jack Haverty via Internet-history wrote: >>>>> >>>>> My immediate reaction was "No, you're not! That's an IMP." >>>> >>>> Welcome to the distinction between a popular perception versus a >>>> professional one. The former lacks nuance, using very coarse metrics. >>>> The latter ought to be more refined, and sometimes is. >>>> >>>> For the world at large, packet switching is really the defining moment. >>>> >>>> For them, the moment is the invention of computer networking, rather >>>> than the invention of linking networks together. (That is, assuming >>>> that are careful enough to avoid confusing Web with Internet...[*]) >>>> >>>> Sometimes, the error is in the later direction. >>>> >>>> Getting even professionals to be careful in talking about email history >>>> is difficult. So it is quite common even in highly technical circles >>>> -- such as a couple of weeks ago for a press release -- for folk to say >>>> that Ray invented email rather than Ray invented networked email. >>>> >>>> d/ >>>> >>>> [*] maybe 20 years ago, taking a Spanish course in Spain, with a class >>>> of much (much) younger folk from all over Europe, the instructor >>>> prompted some discussion in Spanish by asking us about our backgrounds. >>>> I chose to say that I worked on the UCLA networking project, in 1972, >>>> explaining it was the first site on the Internet. One of the very >>>> bright, very young students objected vigorously, saying that the >>>> Internet was invented in 1989.[**] I smiled and tried to explain the >>>> difference but she persisted. The instructor didn't care about the >>>> answer, as long as everyone was talking in Spanish, but this dragged on. >>>> The youngster would not relent. Finally the instructor intervened, >>>> say "Look, you weren't born yet and he was there!" >>>> >>>> [**] I have heard of a name for it, but there should be one that >>>> distinguishes errors that require a lot of knowledge to make. That she >>>> knew of 1989 in the net's history was impressive. My first encounter >>>> with this type of error was while at the University of Delaware, around >>>> 1980, talking to a hotel reservation agent in Toronto. She asked for my >>>> address and when I said Newark, Delaware, she queried "that's a suburb >>>> of Philadelphia, isn't it?" sigh. >>>> -- >>>> Dave Crocker >>>> Brandenburg InternetWorking >>>> bbiw.net >>>> -- >>>> Internet-history mailing list >>>> Internet-history at elists.isoc.org >>>> https://elists.isoc.org/mailman/listinfo/internet-history >>>> >>> -- >>> Internet-history mailing list >>> Internet-history at elists.isoc.org >>> https://elists.isoc.org/mailman/listinfo/internet-history > > >