[ih] Larry Roberts & RD the first electronic mail manager software [was written in TECO on TENEX]

[Steve Crocker]

There were multiple conferencing system efforts in those early years.  BBN
built one that included video.  I moved to ISI in 1974.  The BBN system was
built a bit later; I don't remember the exact date.  I do remember sitting
in a swivel chair, spinning it around, and then watching myself complete
the swivel, so the latency was definitely noticeable.

The Institute for the Future (IFF) developed a conferencing system called
Forum.  It was essentially identical to IRC, except it required everyone to
be logged into the same machine.  A user's input was considered to be a
paragraph.  Even if the user was working at a character-at-a-time terminal,
which most of us were, output would be suppressed until the paragraph of
input was complete.  At that point, any paragraphs from others that had
been queued up were then printed.  If you tried to type new input, you
would not see the echoed characters until all the queued up paragraphs had
spewed out.

This was a remarkably effective and usable system.  I have a vivid memory
of using it to interview a candidate to replace Larry Roberts as head of
IPTO.  On a Sunday afternoon, six of us participated: The candidate, Larry
Roberts, Steve Lukasik, Alex Tachmindji, Bob Kahn, me.  Tachmindji was
Lukasik's deputy.  Bob and I were program managers under Larry.  All of us
except the candidate were based in the DC area.  Some of us were at home;
others were in the office.  The candidate was at home in another part of
the country,elsewhere, working in his den.

I was the most junior person on the call, so I said little but listened a
lot.  Also, Tachmindji had had the least experience with these tools, and I
provided some help to him via a side chat.

I noticed there were sometimes two or even three distinct threads in
progress.  It was not only fairly easy to follow them, it was actually more
interesting than if we had all been in the same room.  We didn't have to
wait for each person to finish talking, so it seemed more efficient than a
regular face to face meeting.

I've often wondered why this mode of interaction is used more frequently.
I've even tried it out when I had control of the group, but the results
weren't great.

Larry left DARPA in late 1973 or perhaps very early 1974, so that pins down
the date moderately well.

Cheers,

Steve


On Tue, Aug 8, 2023 at 3:56 PM Jack Haverty via Internet-history <
internet-history at elists.isoc.org> wrote:

> Just a few years ago, I stumbled across an Annual Report that MIT
> submitted for one year's work in the early 70s.  Since I was there at
> the time, I was curious how history recorded what we were doing then.
> Looking at the section for our group, I found a description of a
> revolutionary implementation of a teleconferencing system that allowed
> people to interact in real time using the ARPANET which had been
> completed that year.
>
> I didn't remember that we had built any teleconferencing system. Of
> course with age comes memory loss.  But I remember lots of stuff we did
> then, but not a "teleconferencing system".   A sign of encroaching
> dementia...?
>
> With further investigation...
>
> A bunch of us at MIT in Licklider's group spent a lot of hours getting
> multi-player MazeWar running on our fancy new Imlac minicomputers.
> Someone added a feature where players could trash-talk each other with a
> shared screen space trying to lure them into an ambush or gloat on
> another kill.   MazeWars of course had nothing to do with whatever
> research we were doing.   Gettings MazeWar going was just a lot of fun.
> We all thought MazeWars was just a cool hack and extremely addictive
> game.   If curious, see
> https://www.digibarn.com/collections/games/xerox-maze-war/index.html
>
> But the experience did reveal, to me at least, the importance of
> latency, and the difficulties of getting a bunch of computers to
> interact over a network.   Imlacs had no I/O except RS232.  So, our
> "LAN" was a star-shaped configuration with Imlac minicomputers connected
> via RS232 to our PDP-10 as the center of the star (7 floors away), and I
> had goosed the RS232 hardware well beyond its spec to achieve almost 100
> kb/sec.  I tried to convince BBN to upgrade the TIP hardware to support
> higher speed "terminals", but was rebuffed -- "The TIP supports
> terminals up to the maximum reasonable speed of 9600 bits/second."
>
> MIT's Annual Report touted Maze as a "teleconferencing system".
>
> Jack
>
>
> --
> Internet-history mailing list
> Internet-history at elists.isoc.org
> https://elists.isoc.org/mailman/listinfo/internet-history
>