ITS as a social medium

[Lars Brinkhoff]

An online discussion on Mastodon triggered long-time ITS and Lisp hacker Kent Pitman to write this insightful blog post:

https://netsettlement.blogspot.com/2024/07/social-computing-before-internet.html

[Eric Swenson]

I tried to post a comment to Kent’s post, but wasn’t allowed to. What I wanted to mention was another very important advantage to the open environment of the early ITS. Because one could trivially find the source for any program, source code was easily studied, and built upon. This applied to MIDAS sources of executables, but also to LISP, MUDDLE, SCHEME, Macsyma, Emacs, and any other subsystem. You found something that interested you, looked over the code, conversed with the developers, experimented with your own version, got permission from the developer to contribute to the code, helped debug and extend, etc. ITS and its environment was a wonderful catalyst for budding software developers. If you didn’t want to collaborate with another developer, you made your own copy with your own enhancements. You told others where your programs were, or put them in standard places, and they got used, appreciated, or passed over. 

There were countless programs developed by “net randoms”, as we called them, who maybe didn’t even have their own home directories, but used the shared USERn directories. They were logged into the ITS machines from all over the world. They learned and they contributed. Some of their programs were adopted inside the LCS and AI Lab.

It was a wonderful, unplanned but successful, social networking experiment. 

— Eric (KC6EJS, E...@MIT-MC.ARPA)

On Jul 22, 2024, at 00:44, Lars Brinkhoff <lars.br...@gmail.com> wrote:



An online discussion on Mastodon triggered long-time ITS and Lisp hacker Kent Pitman to write this insightful blog post:

https://netsettlement.blogspot.com/2024/07/social-computing-before-internet.html

--
You received this message because you are subscribed to the Google Groups "PiDP-10" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-10/71c58d1f-0d33-41ae-809f-f8887b3b1175n%40googlegroups.com.

[Eric Swenson]

I agree. I didn’t find OS all that useful — usually only in response to a LUSER request (program requesting assistance from a registered mentor). But I frequently used COM links (^_C nn) to chat with people and particularly liked QSEND, which, like SEND sent interactive messages, but also worked across machines on the Chaosnet. 

But as you said, one could learn so much by simply being able to list any directory and read any file on ITS (well, except for the directory where the Zork sources were stored on DM!)

Also, the MLDEV device, allowing the normal file related commands and system calls to operate on remote ITS directories and files. If you were, as I was, an MIT-MC user, you could still trivially explore the file systems of MIT-AI and MIT-DM (yes, MIT-ML too, although I rarely paid much attention to that system).

— Eric (KC6EJS)

On Jul 22, 2024, at 01:06, Ken Harrenstien <ice...@gmail.com> wrote:



Such an interesting read.  Thanks for the pointer!

At first I thought it was going to be about the existence of

communications tools, especially ^_C to immediately chat

with anyone who was online (findable by various means,

starting with PEEK).  However, it seems to start with a focus

on the "spying" capability, which I rarely if ever used.  If you

were really curious what someone had been doing, you could

just skim their file directory.

Instead, the most important social aspect for me was the

ability to communicate.  We could either chat directly in

real time (^_C) or :SEND notes (similar to today's "texts"),

also in real time, or send email -- at first fairly crudely.

Because I was deaf (still am :-), this opened up a new world of

communication and community that I never had before,

and I did what I could to make it even better by writing

COMSAT to provide mailing lists and other features both

locally and remotely.  That also turned out to become my

BS thesis, since MIT at the time required undergraduates

to undergo the thesis ordeal; maybe they still do.

Small groups of us also socialized via games, initially

Spacewar on the PDP-6 & 340 display, and later

there would be larger groups (up to 8) heading down to

the 2nd floor very late at night after everyone in DM/CG

had left, in order to play MAZE on their Imlacs.  (Whoa, that

link actually has a lot of great info.)  How interesting

that online multiplayer 3D first-person shooter games

are now one of the primary ways that kids socialize,

even more so during the pandemic.

KMP is right that it was a pretty amazing and productive

environment, one that we didn't fully appreciate at the time

(how could we, not having ventured beyond the walled garden?)

The tragedy of the commons hit us hard when the constantly

increasing number of Arpanet sites caused a constantly

increasing influx of bad apples, although we had this charming

theory that crackers would leave us alone because there was nothing

to be cracked.

Google in the early days was somewhat similar.  If you were a

full-time employee you had access to pretty much everything

inside the freshly-built walled garden that wasn't personal.

Fun times!   Today, of course, you can only enter a few buildings,

source code is tightly restricted, all access is tracked, credentials

must constantly be renewed, 20%-time projects are a

quaint memory, and on and on.  But I've seen enough

viciously bad actors to have boundless appreciation for the

people in our security teams trying to keep us safe.

Sigh... again, thanks for the memories!

--Ken

On Mon, Jul 22, 2024 at 12:00 AM Lars Brinkhoff <la...@nocrew.org> wrote:

An online discussion triggered Kent Pitman to write this blog post:

https://netsettlement.blogspot.com/2024/07/social-computing-before-internet.html

[Lars Brinkhoff]

Kent couldn't reply here, so I'm posting his text per his request.

--MORE--

I would have liked to respond to EJS and KLH, both of whose
comments/viewpoint were interesting. You're welcome to paste these remarks
(this paragraph and the ones that follow) in that venue. They're very
stream-of-consciousness but maybe will provoke additional interesting
counterpoint from the others in the discussion.

I think it can't just be ^_C though because TOPS-20 had these sorts of
tools. I think EJS is closer to it when he points to the ability to read
all sources, but I think it's underappreciated that one could read each
others' sends or mail. I suppose people didn't read each other's mail
much, but sends I think were read really a lot. Something like that has to
be it, because otherwise I could have just initiated conversations via
talk programs on TOPS-20, which I think they had. It is true that ITS had
a lot of variety in that with ^_C, send, mail, untalk, and maybe others.

Something in there made it reach critical mass so that it felt flexibly
social there and lost it in the commercial systems.

Maybe also an "eat your own dog food" thing going on, too. The maintainers
used their own programs, and they were right there to ask questions of.
Everyone was driven by bug reporting, and there was a sense of honor about
really fixing things. Even the most antisocial of them (I'll not embarrass
them by calling them out by name) would sometimes respond to such mail.

Same with Zork. To have been around when it was being designed and to be
able to tell the maintainers some puzzle room was confusing was really
cool because sometimes you'd see a new newspaper in the game with a fix.

There is no substitute for being able to really talk to the maintainers of
a key tool, or as EJS said, to be able to contribute a fix back because so
much was open source. You couldn't do that on a commercial system either,
but that again is something that was recreated in github later.

Lisp Machines inherited some degree of this culture and were somewhere in
an intermediate space. Some had commercial sources, but the availability
of editor commands to find the sources to anything was important.

My recollection is that RMS (Richard Stallman) was adamant about the idea
that if we didn't control the sources to everything from the machine up,
we'd lose control of our world. I didn't agree with him on everything, but
I could see his point on this, even though it was rapidly happening that
the world changed faster and faster and it seemed like we really couldn't
win that race (without forcing the whole world to be open source, which
wasn't really a formal thing back then, and which wasn't likely to happen
even when it became that way).

RMS managed to push that wedge much farther than I ever would have
imagined he could, though. That idea is core to Linux still, even though
Linux is not itself a social system like ITS was, so even there it can't
be all about the sources, but that doesn't mean it's not part of the
puzzle.

Maybe in some sense what was inevitably lost was that ITS was its own
little github and eventually you couldn't house both the community and the
sources on the same machine, or in the same virtual view, or maybe we
haven't gotten to recreating that illusion yet. So we have to go hunting
for copies of sources instead of just using the actual sources. Again that
comes back to my original point about security. It's necessary, with so
many people in the world using things, that we have security now, and
hence copying. We can't all be in the same environment using literally
each others' sources, not even branches. But it was cool that for a little
while we could.