[ih] A revolution in Internet point-of-view - Was Re: Internet analyses (Was Re: IPv8...)
Steve Crocker
steve at shinkuro.com
Sun May 10 19:18:17 PDT 2026
Two points: one small and the other larger.
1. Dave Farber and Dave Crocker would probably like to see CSNET
included in this list.
2. Regarding control of intellectual property, I actively
participated in the [D]ARPA/IPTO computer science research community from
1965 to 1981:
- 1965-66. UCLA Network Computer Linking Experiment. Attempted to
connect the three IBM 7094s on the UCLA campus. Designed and built a
lightweight time-sharing system for IBM 360/40s, intending for them to
serve as front-ends to the 7904s and connect to each other. We read and
understood the Multics papers. Project died, primarily due to political
issues within UCLA. I was the junior member of the team.
- 1966. Worked for Professor Gerry Estrin in the Engineering Dept
when he received the remainder of the project funding mentioned above.
Estrin focused on measuring computer systems.
- 1967 to May 1968. Graduate student in Minsky's AI Lab at MIT.
Participated in documenting Richard Greenblatt's MacHack 6 Chess
program.
Spent some time with Danny Bobrow's AI group at BBN. Gained exposure to
two operating systems for the PDP-10: ITS at the AI Lab and
Tenex at BBN.
And Interlisp at BBN.
- May 1968 to May 1971. UCLA. Intended to spend just the summer but
decided to transfer back to UCLA. Attended the first IPTO
graduate student
conference. 30 people from across the IPTO-sponsored sites,
including Vint
Cerf, Alan Kay, John Warnock, Pat Winston, Bob Balzer and several others
who made noteworthy contributions over the next several years. My first
exposure to the community and the cohesive effect of funding from
DARPA/IPTO.
- Summer 1968. Designed and partially built a time-sharing system
for the XDS, née SDS, Sigma 7 patterned after MIT-AI's ITS. Discovered
Lawrence Livermore Labs already had a time-sharing system for
the Sigma 7.
We asked if we could use it; they readily agreed and cooperated
fully. LLL
is a Department of Energy Lab and was not funded by DARPA but was, of
course funded by the USG.
- August 1968. First meeting of representatives from the first four
Arpanet sites. Primarily graduate students. The Network Working Group
emerged organically, as did the RFCs and eventually the IETF.
- July 1971-August 1974. [D[ARPA/IPTO Research Program Manager.
Primarily focused on AI, Speech Understanding, and Computer Security, but
helped a bit with the Arpanet activities.
- Nov 1974-July 1981. Researcher at USC-ISI. Focused primarily on
Formal Methods (Program Verification.
I share the above so you can assess the following comments.
Until about 1973, I never saw any issues regarding the distribution of
research results. Of course, very few results also had immediate
commercial value, so the question was perhaps moot. However, I have the
strong impression that sharing all kinds of results, including software was
taken for granted across the community and probably beyond.
Toward the end of my time at DARPA, BBN tried to assert control over both
the IMP code and the Tenex code. The IMP code was particularly contentious
because a BBN employee started a competitive company and requested the code.
The contracts supporting these projects had 100% government funding and
gave the government "unlimited rights in data." (I quote these words
because that was the formal terminology in the contract.). As I learned,
those words permit the government to take delivery of the code and do with
it whatever it wants, including distributing it to whomever it chooses.
However, those words alone do not require the contractor to distribute the
code to others.
That incident brought the control of intellectual property arising from the
research contracts to the surface. Until then, the ethic within the
community, as far as I experienced it, was enthusiastic sharing. Indeed it
was a form of validation of the work and hence a mark of success. I could
see the need for a software repository, but I left before I could initiate
action in that direction. In any case, the key point is that until that
time, I believe it was natural and standard practice to share software when
it was of interest. The only "restriction" was appropriate attribution.
When creating the RFCs, we intended the maximum degree of sharing. The
initial rules for RFCs were based on speedy distribution without review or
control. The idea of charging for documents or requiring group membership
was intrinsically antithetical to the ethic of enabling others to benefit
from the work.
Quite a lot changed after that. Free and open source software became a
thing as did a variety of software licenses. When I went to Aerospace in
1981 to set up a computer science research lab, we paid $500 for a Berkeley
Unix license. It was a trivial sum. Nonetheless, someone in the Aerospace
legal department reviewed the license and felt obligated to negotiate with
the University of California Regents. When I finally went to find out what
was taking so long, he explained he was only trying to help me. I asked
him to stop and we were up and running shortly thereafter.
Vint and others can report on how things evolved after from 1974 onward.
Steve
On Mon, May 11, 2026 at 8:49 AM Vint Cerf via Internet-history <
internet-history at elists.isoc.org> wrote:
> "From the earliest days, the ARPANET & Internet were self-funding from
> revenue - yes ARPA & then NSF provided funds for researchers to purchase
> telecom, but then came the regionals (notably NEARnet) that started
> accepting commercial traffic, and then the whole infrastructure was opened
> to the public in 1992."
>
> Miles, that's not quite right. ARPANET was paid for directly by ARPA -
> funcing BBN to build and operate, funding universities and research labs to
> develop host level protocols and applications. The Internet development
> initially consisted of ARPA-funded Arpanet, Packet Radio Net and Packet
> Satellite Net. The came NSFNET, paid for by NSF and implemented by IBM, MCI
> and Merit. NASA Science Internet linking NASA labs. -paid for by NASA.
> ESNET linking DOE leaps, paid for by DOE. The regional networks were
> partially funded by NSF but were required to become self- supporting over a
> 5 year period (initially 3). It was only in 1989 that UUNET, PSINET and
> CERFNET were commercial operators that were formally allowed to connect to
> the USG networks in 1992 (although permission was given to connect
> commercial MCI Mail to NSFNET in 1989 as an experiment undertaken by CNRI.
>
> v
>
>
> On Sun, May 10, 2026 at 8:10 PM Miles Fidelman via Internet-history <
> internet-history at elists.isoc.org> wrote:
>
> > Re. "give away:"
> >
> >
> > 1.
> > Those were the days when, if the government paid for something, the IP
> > went into the public domain.
> > 2.
> > The Internet grew a lot like the airlines & aerospace industry. Between
> > war & airmail, early government spending - for goods & services - paid
> for
> > infrastructure & an industrial ecosystem that then took off on its own.
> > (The phrase "primed the pump" comes to mind.)
> > 3.
> > From the earliest days, the ARPANET & Internet were self-funding from
> > revenue - yes ARPA & then NSF provided funds for researchers to purchase
> > telecom, but then came the regionals (notably NEARnet) that started
> > accepting commercial traffic, and then the whole infrastructure was
> opened
> > to the public in 1992.
> >
> > Miles
> >
> >
> > ________________________________
> > From: Internet-history <internet-history-bounces at elists.isoc.org> on
> > behalf of Jack Haverty via Internet-history <
> > internet-history at elists.isoc.org>
> > Sent: Sunday, May 10, 2026 5:02 PM
> > To: internet-history at elists.isoc.org <internet-history at elists.isoc.org>
> > Subject: Re: [ih] A revolution in Internet point-of-view - Was Re:
> > Internet analyses (Was Re: IPv8...)
> >
> > There was a different climate in the 1960s/1970s. That was the era of
> > the Vietnam War and associated protests and events such as Kent State.
> > Lots of pushback especially from young folks. Even the resignation of a
> > POTUS, under threat of removal from office. Lots of distrust of
> > government. Lots of anger.
> >
> > Is it so different now?
> >
> > But I don't recall that as the dominating factor in the 60s/70s/80s.
> > The USG did not "give away" the Internet. The Internet (and previously
> > the ARPANET) was always considered an Experiment. That followed the
> > charter of ARPA - "Advanced" Research projects Agency. Research
> > produces ideas. Experiments build prototypes to test theories.
> > Research produces knowledge. Freely sharing that knowledge would be
> > beneficial to the "real systems" that followed. Artifacts, such as
> > protocols, algorithms, equipment designs, software, and such concrete
> > items are at best prototypes for the real systems of the future.
> >
> > The "real system" was expected to be OSI. Of course that didn't
> > happen. But it was The Plan. Even the USG had a program called GOSIP
> > (Government OSI Profile?) to plan for the use of OSI throughout USG. As
> > long as The Internet was considered just a research experiment, it had
> > no long-term value and giving it away didn't raise significant
> > objections. It might even help OSI development. OSI was the target.
> >
> > Instead, what happened was an explosion of industry-led alternatives to
> > both The Internet and the OSI promises. In the "multiprotocol router"
> > stage of the late 1980s and 1990s, all sorts of other schemes were
> > produced by companies, each hoping that their technology would be the
> > winner to create the communications infrastructure of the future, and
> > all the others would just fade away. Most of them had some ambition of
> > global domination, e.g., by offering products to create a "global LAN".
> > But the users were impatient, and selected the only technology which was
> > available to them at the time to tie all their IT into a cohesive
> > infrastructure - The Internet and TCP/IP. Its adoption by the USG
> > established confidence that The Internet would live long and prosper.
> >
> > It was straightforward for a corporation to build its own clone of The
> > Internet, separate from but perhaps connected to The Internet for
> > electronic mail service. So lots of corporations did just that. The
> > IT industry noticed, and itself adopted The Internet as its product
> > architecture. A similar history might be told of Unix and Linux,
> > becoming the base IT environment for all those servers on The Internet.
> >
> > So, ... is it so different today?
> >
> > The Internet has clearly won at the levels of datagrams and web. TCP/IP
> > and HTTP and their friends are the de facto standard.
> >
> > But a similar war is now happening at levels where AI and Social Media
> > live, on top of The Internet. I see lots of corporations creating
> > competing AI services or social media platforms, each hoping to be the
> > winner and become the infrastructure of the future. Even electronic
> > mail has fragmented into a blizzard of ways for people to communicate.
> >
> > But unlike the situation 50 or 60 years ago, I haven't noticed any
> > ARPA-like effort to do the research and "give away" the results for
> > others to use. Such things are what The Internet considers "apps" and
> > outside their scope? Perhaps such research organizations exist, other
> > than in corporate walled gardens, and I just haven't been looking hard
> > enough. There may be no technology "waiting in the wings" for users
> > to embrace until one of them becomes the clear winner.
> >
> > Is that an "attitude of society"?
> >
> > /Jack Haverty
> >
> >
> > It seems to me that today's
> >
> > On 5/10/26 11:36, Greg Skinner via Internet-history wrote:
> > > On Apr 29, 2026, at 6:13 AM, Andrew Sullivan via Internet-history <
> > internet-history at elists.isoc.org> wrote:
> > >> Tempted as I am to editorialize about what this might mean for the
> > Internet (I am profoundly depressed about it), I wonder whether those who
> > where involved in the Internet's earliest developments have any
> reflections
> > on the attitudes of the societies at the time. For instance, kc claffy
> > once observed to me that it was an inspired bit of industrial policy that
> > led the USG (partly it seems to me at the prodding of Al Gore, despite
> all
> > the grief he gets about the topic) to give away the Internet rather than
> > lock it into any particular corporate ownership. I know there is another
> > thread that has discussed the BSD-TCP/IP importance, but I guess I'm
> asking
> > for something different: was there a different _social_ environment, in
> > your estimation and upon reflection, than there is (say) today such that
> > the USG could give such a technology away as they did? I find it
> > impossible to imagine that happening today, when every organization
> either
> > public or private seems to be orieted entirely towards maximum short-term
> > financial return on investment, ignoring the longer term benefits. (And,
> > to avoid any doubt, let me be clear that this is not a particular swipe
> at
> > the current USG or any people in charge of it. This has seemed obvious
> to
> > me for a decade or more.)
> > >>
> > > I would also include the Linux importance. [1] [2] Quite a bit of
> TCP/IP
> > became available via open source due to Linux.
> > >
> > > --gregbo
> > >
> > > [1] https://www.oreilly.com/openbook/freedom/
> > > [2] http://www.catb.org/esr/writings/cathedral-bazaar/
> > >
> >
> > --
> > Internet-history mailing list
> > Internet-history at elists.isoc.org
> > https://elists.isoc.org/mailman/listinfo/internet-history
> > -
> > Unsubscribe:
> >
> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
> >
>
>
> --
> After July 1, Please send any postal/overnight deliveries to:
> Vint Cerf
> 1435 Woodhurst Blvd
> McLean, VA 22102
>
> Until July 1, send postal/overnight deliveries to:
> Google, LLC
> 1900 Reston Metro Plaza, 16th Floor
> Reston, VA 20190
> +1 650-224-2788
> --
> Internet-history mailing list
> Internet-history at elists.isoc.org
> https://elists.isoc.org/mailman/listinfo/internet-history
> -
> Unsubscribe:
> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
>
--
Sent by a Verified
sender
More information about the Internet-history
mailing list