Origin of USENET

From zed@mdbs.uucp Fri Aug 10 18:31:03 1990
From: zed@mdbs.uucp (Bill Smith)
Newsgroups: alt.folklore.computers
Subject: origins of news
Date: 7 Aug 90 23:06:20 GMT
Organization: mdbs, Inc.

>From: smb@ulysses.att.com (Steven Bellovin)
>Date: 25 Jul 90 20:06:47 GMT
>Back in the days of 6th Edition UNIX, there was a program called
>``news''.  I'm not sure where it came from, but messages were
>restricted to 512 bytes.  Login was hacked to check if you had
>unread news, and told you to read it.
>When we upgraded to V7, no one wanted to carry forward such an ugly
>program, but some replacement was needed
>Tom Truscott and Jim Ellis, then of Duke's Computer Science Department,
>conceived of the idea of a network news system.  They and I, and a
>few others, got together and designed netnews -- i.e., the network
>news system.  I'll omit a detailed discussion of the history; however,
>we planned to announce it at the Winter '80 USENIX meeting.  USENIX
>had just received its name; it had been called the UNIX Users' Group,
>but AT&T's lawyers didn't like that.  When the new name was announced,
>at the Summer '79 conference, it was met with some hoots of derision --
>the form of the name was felt to be in violation of the spirit, if
>not the letter, of trademark law.  (That is, if one cannot use X in
>a name, why is f(X) legal?)  Nevertheless, a precedent had been set
>within the community, so we used g(f(X)) to generate USENET -- we
>envisioned that it might become the ``official'' network of USENIX.
>As for msgs and news -- I believe that msgs existed at Berkeley at the
>time, but we had neither seen nor heard of it.  I had heard vague
>rumors of the Bell Labs news program (the one that used 14-character
>file names as article titles, and stored them in /usr/news), and in
>fact my very earliest prototype of netnews (a Bourne shell script)
>used the same technique, with one vital difference:  an article could
>be linked into many subdirectories of /usr/spool/news, thereby
>permitting multiple newsgroups.  Subscription lists were in the
>environment variable SUBSCRIBE; the script used the shell's file name
>expansion to construct the actual list.
>			--Steve Bellovin

Wasn't there a precursor to netnews on PLATO?

I almost certain that PLATO is the origin of "talk" and "write".

Bill Smith


%d bloggers like this: