search

PDP-10 / its

Incompatible Timesharing System
Other
834 stars 80 forks source link

ITS pictures #425

Open larsbrinkhoff opened 7 years ago

larsbrinkhoff commented 7 years ago

I have tried to find pictures of ITS machines. I haven't found much.

larsbrinkhoff commented 7 years ago

aap commented 7 years ago

Looks like mathlab had their software running on the AI PDP-6 at the time. Here's another video about project MAC showing the PDP-6 in all its beauty (mirrored for some reason): https://www.csail.mit.edu/videoarchive/history/aifilms/mac-104 EDIT: Not mirrored here: https://www.csail.mit.edu/videoarchive/history/aifilms/mac-45 EDIT: The person at 3:40 here looks like Greenblatt: https://www.csail.mit.edu/videoarchive/history/aifilms/robot-02 Kotok at 5:00 with someone I don't know next to him. EDIT: Assembler code on a PDP-6: http://projects.csail.mit.edu/video/history/aifilms/43-arm.mp4 EDIT: https://www.csail.mit.edu/videoarchive/history/aifilms/worm-55

These AI films are a great find!

larsbrinkhoff commented 7 years ago

There are two pictures of the SAIL machines. The KL10 is in the foreground, and I think I see a KA10 in the background of the second picture. EDIT: A third picture, the KA10. sail-ka10

aap commented 7 years ago

Even if not ITS, here's another photo of Stanford's KA: http://media.csail.mit.edu/picture.php?/3642/category/173

larsbrinkhoff commented 7 years ago

@simhacker, you wrote:

The console of MIT-MC, on the 9th floor Tech Square, with a can of coke. http://www.catalog.com/hopkins/images/mc-console.jpg

Do you still have that picture?

larsbrinkhoff commented 7 years ago

Never mind, found it!

larsbrinkhoff commented 7 years ago

This well may be the only picture of an ITS machine. Please prove me wrong.

SimHacker commented 7 years ago

Here’s where it is now:

http://donhopkins.com/home/catalog/images/mc-console.jpg http://donhopkins.com/home/catalog/images/mc-console.jpg

Also:

The expression on Richard's face is saying, "I don't know, why do you wrap gerbils in duct tape?" http://donhopkins.com/home/catalog/images/jsol-rms-gerbil-liz-mg.jpg http://donhopkins.com/home/catalog/images/jsol-rms-gerbil-liz-mg.jpg

-Don

On 30 Jan 2017, at 03:23, Lars Brinkhoff notifications@github.com wrote:

This well may be the only picture of an ITS machine. Please prove me wrong.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/PDP-10/its/issues/425#issuecomment-276038587, or mute the thread https://github.com/notifications/unsubscribe-auth/ABA17B-J2i6yiBvOvQE3gc456XPKbEWYks5rXcg9gaJpZM4Lj4u-.

larsbrinkhoff commented 7 years ago

Picture of KATIA. Does it look like it's running ITS?

eswenson1 commented 7 years ago

Can't read the screen, even after blowing up as large as I could get.

larsbrinkhoff commented 7 years ago

Could it be the same can of coke?!?

36bit commented 7 years ago

Can't read the screen, even after blowing up as large as I could get.

Yes, more resolution please! :)

larsbrinkhoff commented 7 years ago

I have asked if it's available in higher resolution. Here's the invoice. Lists of serial numbers online says # 175 was in Finland. I suppose it's possible that it was shipped to Stockholm.

EDIT:
Confirmed, Thord Nilsson says it came from Finland:
https://www.tekniskamuseet.se/wp-content/uploads/2017/08/115-thord-nilsson.pdf

eswenson1 commented 7 years ago

MC?

-- Eric

On Feb 13, 2017, at 22:49, Lars Brinkhoff notifications@github.com wrote:

Unknown KL at Tech Square:

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

larsbrinkhoff commented 7 years ago

My mistake, it's probably not Tech Square at all.

eswenson1 commented 7 years ago

Yeah, it doesn't look like the 9th floor at tech square. Besides MC, did MIT have any other KLs?

larsbrinkhoff commented 7 years ago

Noel Chiappa is my go-to guy for these kind of questions, and he saw right away that it wasn't MC. And he said there were no other DEC-10 KLs at the 9th floor, only DEC-20s.

larsbrinkhoff commented 6 years ago

@aap Those video links are now broken.

aap commented 6 years ago

That's unfortunate. Maybe I'll drop them an email. Or you do it, I probably won't have time today.

ams commented 6 years ago

Maybe check archive.org if they got archived?

larsbrinkhoff commented 6 years ago

From CSAIL webmaster:

We are in the process of building a new video archive interface, but in the meantime, you can access the videos from the backup copy of our old server: http://akron.csail.mit.edu/research/videoarchive

EDIT: They moved again. They are now available from https://archive.org/details/AIfilms EDIT: Or https://www.csail.mit.edu/csail_videoarchive

eswenson1 commented 6 years ago

I had definitely heard of the KL-udge addition and I do recall some lights because there was a hack that allowed some pattern to be discerned from the lights. But my memory is playing tricks with me because I’m think the lights were on the bank of hardware to the right of the operator console in that picture of MC. While the 11 was to the left, as can be seen in the picture, there was much more to be seen to the right. Basically big plain cabinets — but I think MC had lights there.

larsbrinkhoff commented 6 years ago

Joel Moses, probably at the MIT-MC console:

larsbrinkhoff commented 6 years ago

Lots of PDP-6 exposure in the Chess section:
http://projects.csail.mit.edu/video/history/aifilms/04-cubechess.mp4

eswenson1 commented 6 years ago

Yes. That’s the right orientation for MC’s LA36 (I think it was) and the 11. And Prof Moses would have used MC.

rcornwell commented 6 years ago

That is a LA36, I tore enough of these appear in college to remember them. Looks like they took off the clear plastic cover over the print area.

larsbrinkhoff commented 6 years ago

Jeff Harris, unknown, and Licklider by an MIT PDP-10 console. But which one? lick-pdp10

larsbrinkhoff commented 6 years ago

Licklider ran the Dynamic Modelling group, right? So it may be DM.

aap commented 6 years ago

MIT-AI (KS10) at LCM (the one on the right, next to the TOAD): http://aap.papnet.eu/pub/img/lcm/vcf_lcm_exhibit/small/P1060410.JPG http://aap.papnet.eu/pub/img/lcm/vcf_lcm_exhibit/small/IMG_20180214_155147715.jpg its unibus cage: http://aap.papnet.eu/pub/img/lcm/vcf_lcm_exhibit/small/IMG_20180214_155245396.jpg and its chaosnet board: http://aap.papnet.eu/pub/img/lcm/vcf_lcm_exhibit/small/IMG_20180214_155343491.jpg http://aap.papnet.eu/pub/img/lcm/vcf_lcm_exhibit/small/IMG_20180214_155404068.jpg (remove "small" in URL for higher res photos)

larsbrinkhoff commented 6 years ago

Thanks! P1060410 isn't Chaosnet though.

aap commented 6 years ago

Copied wrong link by accident. it's correct now.

larsbrinkhoff commented 6 years ago

First picture: Imlac PDS-1 at Dynamic Modelling.
pds1 Second: Imlac and E&S LDS-1 at DM. pds1lds1

jh95468 commented 6 years ago

The picture with Lick is definitely the MIT-DM machine. It was on the 9th floor of 545 Tech Square. The system was arrayed in two long rows. The one in the picture shows the main console. The equipment rack to the left (behind Lick) was the CPU itself, in those very wide cabinets (much bigger than 19-inch). There was a large tape drive (9 or 7 track) to the left of the CPU, and eventually also an Ampex ARM-10 core memory. You can see several micro-tape drives above the CPU and there were some more, to the right of the CPU. Also a TTY controller built by Steve Morton (SGM@MIT-DM) that we used to connect IMLACs top the PDP-10 at 50-1oo kb/sec (standard RS232 was just too slow). The old PDP-6 was also to the right.

Parallel to that row of cabinets and behind were memory cabinets and later several disk drives along the outside wall. The ARPANET IMP was at the right end of that back row. All that equipment had a wonderful view through the large picture windows all along the back wall!

In the picture, you can see a wall off to the left. That was the start of MIT-AI land, with a workshop area behind the wall and computers, robotic arm etc behind that. In the early days, the space between the DM PDP-10 and that wall had the E&S Display, as well as a few terminals. When our group moved and coalesced into our own space on the 2nd floor, the PDP-10 computer stayed behind of course, but the terminals and E&S moved to the 2nd floor. The space vacated then became home to the MIT-MC and/or ML machines.

The 10th (and top) floor was not accessible to the public. That's where all the HVAC equipment was installed. Those computers generated a lot of heat, and had to be close to their AC.

The picture of the "E&S LDS-1 at DynaMod" was taken in the second floor "terminal room". The Imlac pictures were also taken in the second floor. The hacker on the left is Mike Broos (MSB@MIT-DM); can't tell who's on the right or in the second picture.

BTW, we never called it "DynaMod". It was either "Dynamic Modelling" or "DM".

On the 9th floor, to the right of the DM machine, was a wall separating us from the Multics system. Multics had some massive equipment, which apparently required extra support so it didn't fall through the floor to the 8th floor office. So there was a huge steel I-beam running under the false floor between the two rows of DM equipment. Made it very difficult to do things like running cables for memory and I/O busses!

I spent many hours in front of, around, behind, under, and inside those rows of PDP-10 equipment....

/Jack Haverty (JFH@MIT-DM)

larsbrinkhoff commented 6 years ago

Thanks Jack!

It's interesting to compare this floor plan with your description. It was made by @aap from an interview with Greenblatt.

larsbrinkhoff commented 6 years ago

Michael Speciner staring at the 340 display. ms

larsbrinkhoff commented 6 years ago

Finally a picture of the Knight TV!

knight-console

atsampson commented 6 years ago

Ludd's LI, from here via a 2001 post from @larsbrinkhoff (news:853d2qkfk3.fsf@junk.nocrew.org).

image03

larsbrinkhoff commented 6 years ago

Note that there never actually was an LI ITS. MIT got inquiries from Lysator club members, but it never got past that.

But hey, it's not too late!

larsbrinkhoff commented 5 years ago

This isn't ITS, but a relative, MagicSix at the Architecture Machine group.

https://www.youtube.com/watch?v=JgcfI254xEU

It shows heavy use of two Imlac PDS-1, like those used in the Dynamic Modeling group.

At 4:15 there's an especially interesting plot twist. The camera pans in a wide arc across the room. Along the way, we get to see:

jh95468 commented 5 years ago

On 11/16/18 12:26 PM, Lars Brinkhoff wrote:

What is 3C82C0...?

They look like hexadecimal in pairs, representing 8-bit values. Perhaps bus addresses, or something like that, for the cards located behind the plexiglas panel?

/Jack

larsbrinkhoff commented 5 years ago

The AI films moved agin. Now available from here: https://www.csail.mit.edu/csail_videoarchive?f%5B0%5D=video_vocabulary%3A347

Also while looking for this, I found some interviews with AI people:
http://projects.csail.mit.edu/films/aifilms/AAAIfootage/done/

larsbrinkhoff commented 5 years ago

Here's a picture featuring NAMPIC ITS. Is this the ITS logotype? namdrg

larsbrinkhoff commented 5 years ago

Here is a picture from a 1975 brochure from LCS, reconstructed by @pszolovits. Since the accompanying text is about the AI Lab, maybe this is the AI KA10.

EDIT: Comments below say it's more likely ML.

Professor Joel Moses (seated), Associate Director of the Laboratory for Computer Science, and Professor Patrick H. Winston, Director of the Artificial Intelligence Laboratory.

The Artificial Intelligence (AI) Laboratory is physically housed in the same building as the Laboratory for Computer Science. Members of the two laboratories cooperate in several joint projects, e.g., in natural language understanding, in the architecture of a special machine tailored to the language LISP, and in the development and maintenance of common system software.

ai-ka10

larsbrinkhoff commented 5 years ago

Comparing details against DM visible in a previous comment

larsbrinkhoff commented 5 years ago

Noel Chiappa says:

I'm pretty sure that's ML. Its processor was at the far right of a long set of bays.

pszolovits commented 5 years ago

I also think it was the ML console, but I’m not certain.

On Feb 7, 2019, at 8:01 AM, Lars Brinkhoff notifications@github.com wrote:

Noel Chiappa says:

I'm pretty sure that's ML. Its processor was at the far right of a long set of bays.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/PDP-10/its/issues/425#issuecomment-461409645, or mute the thread https://github.com/notifications/unsubscribe-auth/ADQQC5dR5T2VQAH2n3m7AJNfGwIeKfD9ks5vLCOlgaJpZM4Lj4u-.

larsbrinkhoff commented 5 years ago

Thank you, Mr Szolovits!

larsbrinkhoff commented 5 years ago

Here they are again, with a GT40 this time.

moses_winston_gt40

larsbrinkhoff commented 5 years ago

Fabritek memory and some unknown cabinets.

EDIT: I listened to Greenblatt's comments. The one with the large panel is the moby memory. Two the right are DEC core memories. The box on top is for parity.

fabritek Source: https://archive.org/details/AIfilms/81-mathlab.mp4

jh95468 commented 5 years ago

We had Fabritek memory on the DM PDP-10, but it didn't look like that. Cabinets were black. Each memory was about the size of one of those 4 small modules. I don't remember any large piece as shown at the top though. I don't believe this is the DM machine - did the Mathlab system at MIT have Fabritek? We had such problems with them that our later memory was more from DEC and a 256K cabinet from Ampex (ARM-10).