back to article Happy 30th birthday, IETF: The engineers who made the 'net happen

Thirty years ago today, 16 January 1986, the Internet Engineering Task Force – IETF – was born at a meeting in San Diego. It was humble beginnings and the organization that is more responsible than any other for turning a research project into a viable global communications network boasted an initial attendance of just 21 …

  1. Anonymous Coward
    Anonymous Coward

    IETF can stil be functional

    If the working group chairs start saying what I have heard only one of the L2VPN group co-chairs say (quoting off the top of my head, so no guarantee that the quote is exact):

    "Sit down Linda, when you understand it ask the question again".

    Anyone who has been to the IETF know which Linda(s) does this relate to and which company do they work for (the one that has sabotaged standards process in all standards organizations as a policy).

  2. Anonymous Coward
    Anonymous Coward

    Reminds me of a European SPAG*** technical subcommittee meeting on OSI interoperability in the1980s. Representatives of the collaborating companies were constantly having to find anodyne ways to express "3rd party vendor". It was one representative who seemed primarily concerned with protecting his company's business politics - rather than addressing the pure technical considerations.

    At the first meeting we had 15 minutes to come up with a definition of "interoperability". Afterwards the chairwoman told us that another committee had being trying to do that for the last 18 months.

    "Interoperability: The ability to communicate and do useful work"

    ***Standards Promotion and Applications Group

  3. Anonymous Coward
    Anonymous Coward

    Open was an anomaly

    Since most of the new devices will be mobile-like, rather than PC-like, the next big platforms will likely be propriatory protocols, from propriatory apps, distributed from corporate stores.

    Security still won't be a solved problem, and censorship will approach that of 20th century media.

    Hoverboards still won't hover.

    iPhones will be so thin they'll cut dimensional tears if you hold them wrong, but they'll be status symbols only rather than containing real hardware.

    The web will still exist, but you'll have to raise your ship to broadcast depth, and dodge squid-like drones to log in.

    1. DougS Silver badge

      Re: Open was an anomaly

      Proprietary protocols are often built from open ones. i.e. many proprietary protocols are encapsulated in HTTP or XML.

      I don't think the move to less open protocols has anything to do with mobile devices. There were plenty of proprietary protocols in use on PCs in the past - AIM being a perfect example. The reason is that in the past the internet was non-commercial, so there was no incentive for proprietary protocols. Who was going to invest in creating a proprietary email protocol that bests SMTP or IMAP?

      There were only two sources for protocols, the IETF and people who wrote programs to fit a particular need - IRC being a great example of this. Since there was no commercial profit motive, the protocols were open because the source code was open.

      Once the internet became commercialized a profit motive created a reason to keep things closed. Why should AOL open up AIM - keeping it closed gave them millions in additional revenue from people who would have left AOL if they could have used other clients to connect with their friends who used AOL. That's the reason Facebook is closed - if it was an open protocol to connect with their servers, third party clients could be written to block ads aka "sponsored content".

      1. Infostack

        Re: Open was an anomaly

        The settlement-free model has led to more isolation, balkanization and monopoly than people could ever have imagined. But then that's to be expected for inter-networking that lacks price signals that serve not only as important incentives and disincentives to clear supply and demand ex ante, but to share value in a way that promotes rapid promulgation of new technology and universally inexpensive access.

    2. Notas Badoff
      Facepalm

      Re: Open was an anomaly

      (My comment wings on the title whinge)

      A company I worked for early on, went hook, line and sinker for OSI as the next big gotta-have standard and the soon-to-be required network offering. They did a beautiful new implementation of the full 7 layers and all the protocols. I don't think any other company as much (though everyone else said theirs was "just like" the OSI spec) so it was a industry reference and I suppose served as an example that the standard was "ready".

      That hook/sinker turned into a boat anchor. There was no interop as no one else really had an implementation. No 'synergies'. No combinatorial explosion as others adopted the standard. It wasn't SNA. It wasn't TCP/IP. It was fish guts nobody wanted to pass through.

      Regarding standards, you do need to pay attention to the corporate interest (who is sponsoring/pushing the activity and why), while avoiding the worst effects of the corporate interests. Beware the purely academic standard chiefly fostered for purity's sake.

      Over the last 3 decades everytime I saw reference to the OSI 7-layer-model picture I thought "Go for it, you'll have a Ba'al!"

      1. Anonymous Coward
        Thumb Up

        Hold you breath and hope for a third system?

        At least in Symbian a number of "second systems" were built: glorious towers of design, remedying all flaws of the past and anticipating all needs of the future, removing the need for hard user decisions by presenting all possibilities (it slices! it dices! but wait, there's more! and XML too!) If only the pesky world had stopped turning long enough for the company to wade through those mistakes and move on to better third systems we coulda been a contender...

        Oh, and I have already stolen "have a Ba'al" :-)

      2. DougS Silver badge

        OSI model

        That's an interesting example of a standard that never served as a model for real world implementations (other than Notas Badoff's company I guess) but serves only as a model for telling us what layer all the switches, routers, content firewalls and other network products operate in :)

  4. alpine

    RFC

    The clue that explains why it was all so successful is in the initials on all the fundamental standards. They're still called RFCs, or were when I last looked, which I admit was some time ago.. 'Requests for comment'... Cooperation was the key.

    1. Yes Me Silver badge
      Headmaster

      Re: RFC

      Not only "Not All RFCs are Standards" (that's the title of an RFC, btw) but not all RFCs come from the IETF. The one mentioned in the article ("An IETF with Much Diversity and Professional Conduct") is not an IETF document, for example, believe it or not.

      I think the fact that the IETF ecosystem allows for that sort of idiosyncrasy is major clue to its success. And people like Bill Manning who claim to to have gone off elsewhere to do their work are actually still part of the ecosystem, whether they admit it or not.

  5. allthecoolshortnamesweretaken

    It all seems so normal now, like switching on the light or using a car. Anyone under 30/35 years can't remember the pre-internet age. It's good to be reminded that we are experiencing a technical revolution (and it's impact on society) like printing machines, steam engines, electricity... maybe even bigger. Mindboggling.

  6. -v(o.o)v-

    "White Americans only"???

    Wow - I am offended. Even the current chair is Scandinavian. Many chairs where not white and/or American.

    1. Anonymous Coward
      Anonymous Coward

      A more complete quote would have been: "Having been led at the start and for its first 10 years by white American men", which is an accurate description of the IETF chairs.

      From the start (1986) until 2001 the IETF chairs were Michael Corrigan, Phillip Gross, Paul Mockapetris and Fred Baker, all white American men. After that we get some white European men but still white and still men.

      1. Yes Me Silver badge

        still white and still men

        True, but there have been a number of women in the leadership (including IAB Chair, and of course two ISOC CEOs, as well as numerous Area Directors). Of course things can always be improved, and should be. But until companies send 50% women to the IETF, there won't be 50% in the leadership positions. Ditto for all other aspects of diversity.

  7. Anonymous Coward
    Anonymous Coward

    RFCS and a politics

    Back in the day, we had two competing standards. TCP/IP and a rough and ready 'let's just make it work' system developed by stepwise refined RFCS and Interop and a top down bureaucratic system of protocols designed by technocrats to perpetuate the hegemony of big telecoms carries, called X25 and friends.

    Having failed to get their system widely adopted - back then it was widely accepted that the protocol stack to run it properly would exceed the entire RAM of the current machines - the bureaucrats and purveyors of Big Government have set about trying to cripple the Internet by imposing regulation on it, as they are busy crippling Europe and the West, by imposing regulation on it.

    The people with the ultimate power today are the sort of engineers who created the IETF and the Internet. The sort of people who want ultimate power today are the suits who control IBM, BT, C & W and so on, who with their chums in Brussels and the White House, under the guise of 'liberal socialism' seek to restrict personal freedom of everyone and everything.

    I urge readers to ask themselves one question. Which has affected society more, in a positive way ? Marxism, or the Internet?:

    1. Mike Shepherd

      Re: RFCS and a politics

      The words "take a couple of aspirins and lie down a while" come to mind.

      1. Yes Me Silver badge

        Re: RFCS and a politics

        If you want a good read while lying down with those aspirins, try "23 Things They Don't Tell You About Capitalism" by Prof. Ha-Joon Chang. (Yes, it's directly relevant to why AC's rant is rubbish.)

  8. Tom 7 Silver badge
    FAIL

    IETF minutes in PDF?????

    They'll get the hang of it one day!

    1. Yes Me Silver badge

      Re: IETF minutes in PDF?????

      Minutes from 1986 in PDF ... think about it ...

      They switched to HTML in early 1994 (http://www.ietf.org/proceedings/29/)

  9. jake Silver badge

    Uh ...

    I believe TCP/IP went live on 01/01/83 ...

    I helped with NCP at LBL/ Stanford / Ames / Tymshare / Moffett / Sumex (early SAIL? I can't remember ... I must be getting old), SRI, Xerox, Rand, UCLA et alia [1975-1978] while pursuing an engineering degree at Berkeley (thaanks, ken!) ... this lead to another degree participating in developing TCP/IP with Cerf & Co. at Stanford, which in turn lead to Flag Day, 01/01/83, when the NCP to TCP/IP switch-over took place.

    TehIntraWebTubes was already firmly in place long before the IETF existed. Or thought of.

    1. Pat Att

      Re: Uh ...

      And all that while planning your first Challenger space shuttle mission as Commander. Great work!

      1. jake Silver badge

        @Pat Att (was:Re: Uh ...)

        I had nothing to do with any of the the OV-099 missions. Nor have I ever claimed otherwise. Why do you want to put words into my mouth/keybr0ad?

        Might want to look within, Pat.

        HAND

    2. Anonymous Coward
      Anonymous Coward

      Re: Uh ...

      I was going to comment on here about how this article hadn't mentioned Jake, as I thought he created the Internet. It turns out he did and these guys were just coat-tailing off his work.

      Phew, world order resumed...

      1. DougS Silver badge

        jake is Al Gore?

        Who knew?

        1. jake Silver badge

          Re: jake is Al Gore?

          No, DougS. I'm not Al Gore.

          Al Gore takes his personal large four-engined passenger jet all around the world, to proclaim to any idiot who will listen how green he is.

          Me, I hitch up the Percheron to the buck-board to go into town.

      2. jake Silver badge

        @AC 22hrs (whatever that means,ElReg) (was:Re: Uh ...)

        Please, do try to read for context. I never claimed to invent TehIntraWebTubes. I WAS, however, an engineering student, involved in computers and networking, at Berkeley and Stanford during the mid '70s and mid '80s.

        You personally may have never been in the wrong place and wrong time ... but some of us (obviously) were, or this conversation wouldn't be taking place.

    3. Michael Wojcik Silver badge

      Re: Uh ...

      TehIntraWebTubes was already firmly in place long before the IETF existed. Or thought of.

      Yup. I know it's de rigueur for the Commentarati to downvote jake and complain that he's claiming he's done everything, but his history in this post is spot-on (and I don't see any overweening claims in it, either).

      As jake says, the TCP/IP Internet was already three years and change old when the IETF was formed, and its precursors (the NCP Internet, ARPANET, etc) quite a bit older. ARPANET went live in 1969.

      Similarly, the RFC system was invented in '69 by Steve Crocker.

      The IETF was, and remains, hugely important in the evolution of the Internet, but it was indeed formed a good while after the Internet came into being. I don't think the article claims otherwise, but people unfamiliar with the history might be confused.

  10. richardcox13

    > the internet grew faster than any technology has ever grown in the history of man and yet it never failed, faltered or fell over

    It may never have failed but it has certainly faltered.

    I can't be the only one who remembers the "Great Internet Worm" of 1988, when some 40% of the hosts online were taken out.

    1. jake Silver badge

      @richardcox13

      In 1988, TehIntraWebTubes wasn't exactly ready for PrimeTime.

      In 1988, the Morris Worm affected the Sun3 systems at work. It did NOT affect my personal DEC system under Bryant Street in Palo Alto. Why not? Because I didn't really trust remotely available software being made available to all and sundry, and had all that stuff turned off on the internet-facing gear. In modern terminology, I was using the DEC kit as an early version of what we now would call a "stateful firewall" (behind it was an AT&T PC7300 "UNIX PC", running the actual server code).

      I had warned my company of the potential vulnerability. TCP/IP wasn't perfect, was still a research platform, and those of us in the trenches knew it.

      As a side-note, TCP/IP is STILL an imperfect research platform. My mind absolutely boggles at the number of international corporations (and governments!) who assume it's invulnerable!

      1. martinusher Silver badge

        Re: @richardcox13

        This is the power of Marketing, again. I don't know that many people who understand TCP, they just know its 'reliable' so trying to root it out from unsuitable applications is virtually impossible. My particular beef is people who take a serial protocol, typically a home-made serial protocol derived from a ping/pong terminal exchange, and put it on TCP. Trying to explain to them that you can't put framed traffic on a stream protocol without a framing mechanism .... well, its just too difficult, and anyway, what do I know? The basic idea was sound but I think people forgot that that it wasn't a piece of software, it was a pair of machines with forward and feedback path delays -- it works fine on an instantaneous network but once you start working in the real world...

        (I've found the most effective way to deal with these issues is retirement.)

  11. martinusher Silver badge

    Remember OSI?

    The IETF were just a bunch of anarchic engineers. If you want to engineer a proper network you need a top down standards driven approach. Hence OSI, aka "A large bag of ordure". It did everything ass-backwards (including bit ordering...). It endures, though -- you can't see a presentation about networking of some sort without seeing a slide of the "Seven Layer Model". It is still infesting organizations like the IEEE (who insist on running OSI Type 2 traffic as the basis for running WiFI )(for example).

    The IETF may be suffering from corporate/political takeover but its nowhere near as bad as having it built-in from Day One.

    (BTW -- I watch the "Internet of Things" gyrations with interest. Its not as if connecting 'things' to a network is a novelty, we haven't done a whole lot of it in the past because its been pointless, there just wasn't enough information to exchange to make the engineering cost worthwhile. Still, I suppose its the Next Big Thing so let's find the bandwaggon.....)

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019