Home page logo

interesting-people logo Interesting People mailing list archives

Re: NSF and the Birth of the Internet LAN Interconnects incl (DCS Token Ring)
From: David Farber <dave () farber net>
Date: Wed, 20 Aug 2008 10:15:53 -0400


Of course you are right -- memories are fragile. We developed the DCS Token Ring as part of the NSF DCS (Distributed Computer System) effort at UCI Irvine. It was the first operational ring and was fundamental part of the DCS computer system. As part of an Arpa effort we transfered the design of the working rings to MIT for use in their activities. Basically they worked with the DCS Ring stripped of the process addressing mechanism we designed in for use in supporting a distributed high availability multi-machine system.

BTW We also spent a fail amount of time with IBM engineers who took the basic architecture of the DCS Ring and produced the IBM Token Ring. For several years they declined to recognize the source of their design but eventually acknowledged the source.

BBTW The DCS system was , to my knowledge, the first OPERATIONAL high availability fault tolerant computer running a fully distributed operating system -- vintage 1973 ish.


Begin forwarded message:

From: "David P. Reed" <dpreed () reed com>
Date: August 19, 2008 9:15:56 PM EDT
To: dave () farber net
Cc: ip <ip () v2 listbox com>
Subject: Re: [IP] NSF and the Birth of the Internet LAN Interconnects incl (DCS Token Ring)

Don't trust memories from pinging for "history". As you know, Dave, the "token ring" partnership cited below was with MIT - specifically Saltzer, Pogran, Clark and yours truly.

In fact, because I devoted a bunch of my time to developing LAN- centric protocol architectures in those days,Dave Clark, Ken Pogran and I wrote an "invited paper" for Proc. IEEE called "An Introduction to Local Area Networks" that was the first survey paper that covered hardware and software architectures from a variety of creative groups - including your group, Cambridge U, Xerox PARC, etc.

[Side note: This work on LAN protocols was my debut into the "internetworking" (small "i") arena - since LANs by themselves are not terribly exciting, but they also introduced a style of communications that is not about "remote resource access". Of course, there was an inclusive, interconnected community in LANs as well - it would be hard to name the "father of LANs" other than to say it wasn't me. My paper with Saltzer on "source routing" cited your group's work and built on it as well as the work of PARC and Cambridge U., for example.]

David Farber wrote:

Begin forwarded message:

From: Miles Fidelman <mfidelman () meetinghouse net>
Date: August 19, 2008 5:38:16 PM EDT
To: dave () farber net
Subject: Re: [IP] Re:  NSF and the Birth of the Internet

A few tidbits I obtained by pinging some of my former BBN colleagues, who's memories of those days are a bit clearer than mine, and they pointed to some very early LAN interconnections, including:

- the BBN "Fibernet" and "PTIP" networks


- Xerox PARC (obviously)

- a token ring network that was a collaboration between BBN and "Dave Farber's group at UC Irvine"

- a MITRE network that was based on RF over their institutional CATV network (remember "broadband ethernet"?)

- and an interesting tidbit that the old ARPANET IMPs, in their X.25 flavor, "had a facility that enabled an X.25-connected host to exchange IP traffic with an 1822-connected host

All of these were in the 1970s.

Re. Gopher: maybe a pain, but I ran a few rather large gopher sites back in the day. It sure beat managing files by hand. This included a little exercise we (my old Center for Civic Networking) conducted for the Federal Trade Commission. They were doing a rulemaking on consumer protection for online transactions (in those days, focusing on Compuserve and AOL), and we convinced them that they really needed to reach a larger audience than the DC lobbyists - otherwise they'd end up with rules that were onerous for all the small business folks who were just starting to find larger markets via network. We ended up OCRng every comment that came in (no allowance for electronic submission in those days), and putting them up on a gopher server. It was amazing how many people accessed the site, and the volume of comments that came in - and the resulting rules ended up pretty reasonable. As far as I can tell, this is the first time any portion of a Federal rulemaking was put online.

We were also automatically gatewaying some email lists into gopher- supported archives. Also beat managing things by hand (before easy- to-use archiving software).

We hosted with a little hosting outfit on Sherman Street who was providing gopher support for some very large publishers (I should remember names and details, but they escape me). They shared the building with O'Reilly publishing.



Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com

  By Date           By Thread  

Current thread:
[ Nmap | Sec Tools | Mailing Lists | Site News | About/Contact | Advertising | Privacy ]