[ih] Fw: Cisco origins (Was: when did APRANET -TIPs become known as -TACs)

[Barbara Denny]

 Tony is great to correct me (I certainly wasn't considering any time at a house).  I think the parking lot episode was the O'Brien location.  I thought it was in East Palo Alto. I never know where the borders are when I am driving around.
barbara
   ----- Forwarded Message ----- From: Tony Li <tony.li at tony.li>To: barbara denny <b_a_denny at yahoo.com>Cc: Brian E Carpenter via Internet-history <internet-history at elists.isoc.org>Sent: Monday, September 29, 2025 at 12:51:19 PM PDTSubject: Re: [ih] Cisco origins (Was: when did APRANET -TIPs become known as -TACs)
 
Hi Barbara,

Again, I can’t post to I-H, so plesae feel free to forward.

The original cisco office was Len and Sandy’s house in Menlo Park.

The first real office building was at 1360 WIllow Rd., also in Menlo Park and strictly speaking not in EPA.  From there, the company moved to O’Brian Dr. in Menlo Park, even closer to EPA.  Gunfire was common in the neighborhood and at the time EPA was the per-capita murder capital of the US.

One January, we came back from the New Year’s holiday to find a hole in a skylight and a slug on the carpet.

We were very glad to leave that neighborhood.

T


> On Sep 29, 2025, at 12:39 PM, Barbara Denny via Internet-history <internet-history at elists.isoc.org> wrote:
> 
> The original Cisco office was in a building in East Palo Alto.  For those who don't know about the Bay Area, East Palo Alto was not a very great neighborhood when the company started.  I remember mentioning to Greg once that I thought I saw bullet casings in the parking lot.  I don't think he was surprised.
> barbara
>    On Monday, September 29, 2025 at 09:24:56 AM PDT, Craig Partridge via Internet-history <internet-history at elists.isoc.org> wrote:  
> 
> Like Guy and Noel, I'll try to be careful here.  Like Guy, I spent time in
> the first cisco office (2nd floor of a building though my recollection is
> that it was actually in East Palo Alto by a few feet).
> 
> The "SUN" machine on which the first cisco routers were based was a
> Motorola 68000 CPU on a Multibus chassis.  A notable aspect of cisco's
> early days was their willingness to (1) support any link that you could get
> a Multibus card for (I think Greg Satz spent his first two years just
> writing new drivers :-)); and (2) a willingness to deal with repackaging
> what was a fairly common architecture in any way a paying customer wanted
> -- I remember Len, c. 1988, proudly showing off a router in a hard-sided
> briefcase they'd built for a customer.  In short, aggressively "connect
> anything, any way you want".
> 
> Another note, based on a conversation I had with Dan Lynch at the time.
> Many 1980s Silicon Valley startups that grew swiftly had CEO challenges as
> they hit various logistical/organization challenges that come with scaling
> to larger cash flows and employee bases.  Cisco was one of the first
> companies to decide to address that problem by having CEOs for stages: I
> don't know who picked Bill Graves as the first CEO, but Dan told me c. 1991
> that Morgridge  (CEO as of 1988) was brought in to take cisco to a certain
> size and that Chamber's hiring in 1990 was to put in place the person to
> take over once that size was reached and, indeed, Chambers became CEO in
> 1995.
> 
> Craig
> 
> On Mon, Sep 29, 2025 at 6:50 AM Guy Almes via Internet-history <
> internet-history at elists.isoc.org> wrote:
> 
>> Noel,
>>    There are many on this list that were in/about Stanford during the
>> mid/late 1980s, so I'll try to be cautious.
>>    It should first be noted that cisco (originally with no capital C)
>> was just Len and Sandy doing whatever consulting/whatever they were
>> doing while also working at Stanford.  I believe they had a PDP-10ish
>> (TOPS-20?) system at home to use in that consulting work.
>>    You allude to Len's work supporting systems at Stanford.
>>    I believe that he (at least eventually) also worked on the routers
>> (then usually called gateways) being built as part of the Stanford
>> University Network (i.e., SUN) project.  As hardware, the SUN gateway
>> was basically identical to the original SUN workstation (both built as
>> part of that SUN project).
>>    The story of how the workstations were "spun off" into Sun
>> Microsystems and the gateways were "spun off" into Cisco is complicated
>> and worth understanding.
>>    My first in-person visit to cisco was in early 1987, by which time
>> cisco was a handful of people working out of rented space on the 2nd
>> floor of an office building in Menlo Park.
>>    It was so small.  Their "board room" had a table, a few chairs, and a
>> map of the world.
>>    Len had evidently thoroughly received your message about the router
>> business being important and he was all in.
>> 
>>    I believe that the original cisco gateway/router was essentially the
>> SUN gateway with some additional cisco code.  I suspect that the
>> multiprotocol part and the terminal concentrator part came for free with
>> the SUN gateway code base.
>>    While it is true that cisco benefited from that SUN gateway code
>> base, it's also true that the cisco folks worked very hard.  Even within
>> the IP router functionality, they quickly supported their proprietary
>> IGRP along with RIP and the Hello Protocol (from Dave Mills' Fuzzball
>> gateway project which was used in the 56-kbps prototype NSFnet backbone).
>> 
>>    Also, do you remember when the visit to Stanford that you described
>> here took place?
>>    For all its messiness, the history of how cisco was in a position to
>> produce its routers by 1987 was important to how several of the the
>> NSFnet regional networks and other late-1980s Internet components grew.
>> 
>>    Regards,
>>        -- Guy
>> 
>> On 9/29/25 5:15 AM, Noel Chiappa via Internet-history wrote:
>>> 
>>>      > From: Barbara Denny
>>> 
>>>      > BTW I am pretty sure Cisco's first product was not a router. I
>> have
>>>      > seen websites only talk about routers in the history of Cisco.
>>> 
>>> This is from memory, so take it with a big grain of salt. (Not iterested
>> in
>>> researching the point.) But I was very closely associated with these
>> events...
>>> 
>>> My memory is that they did, roughly simultaneously, a multi-protocol
>> router
>>> _and_ a terminal concentrator - in fact, they were more focused on the
>>> terminal concentrator at the very start. The router was based on prior
>> work
>>> at Stanford - Bill Yeager's work. I don't know about the Cisco terminal
>>> concentrator - although I retain a vague memory that it had its roots in
>>> prior Stanford work too. (I did a Web search for "Cisco terminal
>>> concentrator origins", but nothing turned up.)
>>> 
>>> 
>>> Amusing (in retrospect) story about this: Yeager's boxes were used
>> _inside_
>>> Stanford - but they never had ARPANET support. Stanford's first ARPANET
>>> gateway was - a 'C Gateway' from MIT! So I was out there, sitting in the
>>> terminal room in Margaret Jacks hall, working on 'Golden' (their C
>> Gateway),
>>> and in walks Len Bosack - who was then running Stanford's timesharing
>> system
>>> (a TOPS-20, IIRC).
>>> 
>>> We fell to chatting, and I explained to him my insight into why there was
>>> going to be a _huge_ market for routers (roughly fixed ratio of
>> PC's/routers;
>>> common projections of how large the PC market was going to be; A+B=$$$.)
>> A
>>> year or so later, this compny called Cisco appeared.. :-)
>>> 
>>> I still have the configuration files for 'Golden'! (The binary loads for
>> it
>>> had to be created at MIT - at least at the start. I don't know if that
>> ever
>>> got moved to Stanford.)
>>> 
>>> A long time ago, in a universe far, far away...
>>> 
>>>      Noel