Open larsbrinkhoff opened 7 years ago
Theoretically I could install ITS on my KS10 and make it two.
Well, that is, if we're only talking about hardware. There are a few ITS machines up 24x7 running under emulators and hosted on the internet. I, Björn, and Paul have three of them.
Yeah, ITS is pretty much run by Swedes now. :-)
RFC 419 pinpoints the time when DM got a pager:
on or about December 26, 1972, the system will cease operation for a period of at least 10 days and possibly as long as three weeks. During this period, we will be installing a paging system.
@mgthompson @philbudne I found a DEC invoice for the Stacken KA10. I believe the serial number is 175.
See page 31 here: http://elvira.stacken.kth.se/~mz/sp/stackpointer-1985-6.pdf
Dave Conroy's machine is called TX:
http://fpgaretrocomputing.org/pdp10x/
According to reliable sources, the Stacken KA10 probably never ran ITS. There was a KS10 ITS, though.
KA ITS at Stacken?
Date: Sat, 3 Oct 87 23:02:14 EDT From: Alan Bawden Subject: The operating system that wouldn't die! AAAAIIIIEEEEEE!!!!! To: INFO-ITS@AI.AI.MIT.EDU
I thought I would take this opportunity to spread the word about something that I don't think has been very widely publicized. Some of you may recall that a while ago some fellows in Sweden contacted us about running ITS on various PDP-10's that they owned? Well, we mailed them a set of tapes for bringing up ITS on their 2020, which they were able to do without too much trouble (their biggest problem was -my- fault). That all happened over a year ago. Recently we learned that these guys have successfully -built- ITS paging hardware for their KA-10, and have ITS up and running there as well! Totally Amazing.
I thought the ITS KA hardware had gone to sweden, and I assumed the paging boxes (and whatever remained of the PDP-6 did too)....
There's a message copied into a comment here which says the KA AI went to the MIT Concourse undergrad program. I haven't seen anything about ML or DM. The KL MC indeed went to Sweden.
Yeah, the one that MIT gave to Stacken was the MIT-MC KL-10A.
As I think I recall the story: the Stacken KA-10 was one they found on their own (on a loading dock somewhere? dunno). MIT's introduction to the Stacken folks was them writing to us saying they'd heard we had a free OS which would run on that hardware and could they please have a copy. We said sure, but it requires custom paging hardware, to which they replied by asking for copies of the paging hardware blueprints along with the tapes. A few weeks later they reported that they'd built the hardware and had it running -- after correcting a bug in the blueprints, so they sent us the fix.
Hello SRA,
Thanks for the confirmation! I have talked to Stacken people, but they don't remember anything about a KA ITS, just KS.
They bought their first KA10 from DEC for less than $1. There's a copy of the invoice somewhere. It was pre owned and came from Finland.
Pandora Berman answered a question from me:
I see no record of what happened to the KA ML and DM. Do you know?
To the best of my knowledge, they were both scrapped. We have the ML front panel in our home office; Alan thinks an old acquaintance of ours has the DM paging panel. Alan says that for the last few years of its life, the ML-KA was in increasingly bad shape -- down more & more of the time. Alan used it to edit his master's thesis, & was essentially the only ML user for that summer; the previous principal user group, the Med. Group folks, had all evacuated to the farm of Vaxen. Shortly after Alan finished his thesis, ML became too unreliable even for text editing. We think the same sort of thing happened to DM.
ML and DM:
Leonard H. Tower wrote:
Noel Chappa wrote:
They were in the basement of 545 TS, in a random storage room that MIT owned, for a while. I seem to recall that at some later point, when they needed the space, they were taken out and moved somewhere else (not sure whether it was off-site storage, or the scrap bin).
Too many years ago, I saw several racks of DM including the front panel at the old MIT Surplus Property Depot out on the west end of Albany Street. I remember my amazement that it was there, and hadn't been salvaged for a better fate. Don't know what happened to DM after that.
MIT Arpanet hosts:
IMP 6
IMP 44
IMP 77
Network map for some part of MIT, from 1987:
http://www.domain-logic.com/images/1987-network-map.gif
I'm pretty sure that XX's Chaosnet interface was on subnet 1 or 6. Well, OK, it's more confusing than that.
XX had three network interfaces, two of which only spoke IP, the third of which only spoke Chaosnet.
XX's real Chaosnet interface was a direct link to XI (XX's Chaosnet front end PDP-11), but, per Bede's note, that link is not shown. I'm pretty sure that XI had an original Chaosnet hardware network card, not one of those newfangled Ethernet interfaces, so it would have been connected to a physical Chaosnet subnet.
XX did also have an Ethernet interface, which, well, and here's where I think the error crept in, was on subnet 26, not subnet 26. See, in those days the MIT custom was to number subnets in octal. So XX's internal IP address of 18.26.0.36 was on subnet 32 (as shown).
Thanks @sraustein. We have the full configuration for all MINITS routers, including XI:
https://github.com/PDP-10/minits/blob/master/mits_s/config.849#L708-L768
It looks like the CH11 device is commented out, and there are two Interlan interfaces one of which was unreliable.
Another network map, just some Lisp machines, AI, MC, SPEECH TOPS-20, and a MINITS terminal concentrator.
Old HOSTS.TXT files has this to say about the ITS machines and the upcoming Twenexes:
Date | DM | AI | ML | MC | XX | OZ |
---|---|---|---|---|---|---|
1983-03-05 | 1/6 | 2/6 | 3/6 | 3/44 | 0/44 | 3/77 |
1983-05-27 | 10.1.0.6 | MIT-AI-RESERVED | 10.3.0.6 | 10.3.0.44 | 10.0.0.44 | 10.3.0.77 "MIT-AI" |
1983-06-15 | 10.1.0.6 | MIT-AI-RESERVED | 10.3.0.6 | 10.3.0.44 | 10.0.0.44 | 10.3.0.77 "MIT-AI" |
1983-11-04 | Commented | MIT-AI-RESERVED | 10.3.0.6 | 10.3.0.44 "MIT-DMS" | 10.0.0.44 | Commented |
So it seems AI was shut down around April 1983, and DM some time fall 1983. No indication when ML went down; it's removed between 1985-03-03 and 1986-02-05.
DM KA10
@dlebling wrote in a note to his "MAC-beth" poem (see #1336):
The old project MAC Dynamic Modelling Group got its PDP-10 in 1971. It only had 32K of memory, and the group spent a lot of time building more. It was considered a major achievement when there was first 96K of memory on the DM machine.
There was no memory, and no paging hardware. The system (a heavily modified version of ITS) gave you a contiguous chunk of core, and it was yours until you gave it back: no swapping out, but no sharing either!
DM was finally retired in late 1983.
DM ended up with 512K of memory: the original MA-10s, along with an MD-10 (I still have some spares for that), MF-10, and finally a 256K Ampex ARM-10 ('75 or so). There was some Fabritek memory that never worked well. It had the usual SC-10 paging box, from sometime well before 1974. IIRC, it went from 5 RP02 drives to 3 RP02s and 3 RP03s about the same time the ARM-10 was added. This description sounds more like the DM PDP-6, which I never saw powered up.
There is a SYSTEM; CONFIG 17 file from 1975. At this point DM had:
There's this description from ARPANET. The date "December 1978" is mentioned. The hardware section lists:
What was the serial number of this machine?
On Thu, Oct 11, 2018 at 6:52 AM Lars Brinkhoff notifications@github.com wrote:
There is a SYSTEM; CONFIG 17 file from 1975. At this point DM had:
- 512K memory.
- DC10 with 6 RP02 disks.
- One TM10B.
- 8 TTYs connected to the TK10, and 32 to the Morton box.
- E&S LDS-1 display.
There's this description from ARPANET http://pdp-10.trailing-edge.com/mit_emacs_170_teco_1220/01/info/mit-dm.txt.html. The date "December 1978" is mentioned. The hardware section lists:
- 512K core.
- A PDP-11 with 56K core.
- 3 RP02, 3 PR03, and 1 RK05 (for the 11?).
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/PDP-10/its/issues/181#issuecomment-428910776, or mute the thread https://github.com/notifications/unsubscribe-auth/AEFUHtjz1wAtLoWA0VrvGrhoSgES3UhDks5ujyLWgaJpZM4LHmwI .
-- Michael Thompson
As per your list: "144 TOPS-10 MIT-DMS, MA KA10". Well, TOPS-10 isn't right. There's another line "?? ITS MIT-DMS Cambridge, MA KA10".
Also, you have MC listed as a 1088. Wouldn't that be a dual CPU? I have seen 1080.
What we don't know are the numbers for the AI PDP-10, and the DM PDP-6.
As per your list: "144 TOPS-10 MIT-DMS, MA KA10". Well, TOPS-10 isn't right. There's another line "?? ITS MIT-DMS Cambridge, MA KA10".
I updated this.
Also, you have MC listed as a 1088. Wouldn't that be a dual CPU? I have seen 1080.
Fixed
What we don't know are the numbers for the AI PDP-10, and the DM PDP-6.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/PDP-10/its/issues/181#issuecomment-428917613, or mute the thread https://github.com/notifications/unsubscribe-auth/AEFUHoppMxLVjhjWRgpvvSlNo7hXv7c_ks5ujykVgaJpZM4LHmwI .
-- Michael Thompson
DM the machine never had an 11 attached. What's mentioned was probably a standalone machine used on the Morse code project DM the group was doing starting in '75.
@jh95468 wrote elsewhere about the Morse code project and their PDP-11:
During my tenure at MIT-DM (1970 through late 1977), the only PDP-11 I recall attached to the DM machine was an 11/05 or /10 that we used as a signal processing front-end for the "Hand Sent Morse Understanding" project. That PDP-11 was in a small lab on the 2nd floor, and connected to the PDP-10 through the same coax lines we used to connect the Imlacs. So it almost certainly went through the "Morton box". The PDP11 interfaced via RS232 to a Collins radio in the lab as well as some custom hardware we built (I remember spending many hours wire-wrapping and debugging augat boards). The PDP-11 did things like A-D conversion and then FFTs of incoming signals, and the custom hardware contained PLLs to lock on to and track individual Morse signals.
SYSTEM; CONFIG 39 to 81 (1976-1980) includes DEFOPT NTYP==1 ;A REAL KLUDGE TO TALK TO THE 11
.
You wrote:
There is a SYSTEM; CONFIG 17 file from 1975. At this point DM had:
* 512K memory.
* DC10 with 6 RP02 disks.
* One TM10B.
* 8 TTYs connected to the TK10, and 32 to the Morton box.
I think the Morton TTY controller supported a maximum of 24 lines. You might want to double check the numbers.
There is no way ITS with 512K memory could support 40 active users.
I don't recall the Morse Code project ever having an 11. It had a fancy digitally controllable radio, but that was on the 2nd floor.
In DM's early days, there was a PDP-9 right next to it (on the left, for those who recall the lab layout). It was never "attached" though. The 9 was removed when DM got additional disk drives and the space was needed.
The DM PDP-6 was on the right side of the line of cabinets. It was on one of the DM memory ports, so you could theoretically copy data to it. The E&S worked the same way. I don't recall the 6 even being powered up except in fun ("I wonder if it still works?") during the time I was there. When it was finally junked I helped take it apart and unplug it from the 10. This was, IIRC, when the MIT-XX TOPS-20 machine was coming and the space was needed.
Thank you @david-moon. That is indeed something to look into. These are the few facts I can find:
20
, which is 16. I had this wrong above.33
, i.e. 27.Unfortunately, I don't have old versions of SYSTEM; TTYTYP which would have listed all terminals.
"There is no way ITS with 512K memory could support 40 active users."
At times it couldn't support 4 active users. MUDDLE compilations and ZORK instances could be huge. It might support a lot of inactive users, though.
More about tight memory on the early DM ITS, from @dlebling618's MAC-beth email:
occasionally you would hear "I need to do a MIDAS." "Okay, I'll kill my DDT." MIDAS needed fifteen whole pages of memory! It was considered giant. MUDDLE needed 32, and was practically beyond the pale; true overconsumption to use it.
There was a program called GOBBLE at one point which you ran by saying :GOBBLE TECO (for example). It knew TECO was 6 blocks, so it sat around trying to get 6 blocks, and when it did, it .VALUEd ":KILL
:TECO ". It was exciting watching it print out how many blocks it had gotten.
Date: Sat, 8 Feb 86 03:15:22 EST From: "Pandora B. Berman" [CENT%AI.AI.MIT.EDU@MC.LCS.MIT.EDU] Subject: Reports of ITS death have been highly exaggerated
the semi-original AI (that is, the KA-10 rather than the PDP-6) was flushed 3 years ago. the hardware was given to a bunch of hackers from Concourse (an MIT alternative undergrad program)
Concourse is a first year "alternative undergrad program", basically about 50 students are taught together in the conventional way, except as a group and in a more integrated fashion than normal, and the Institute likes it if for no other reason than that it costs them about half as much as the normal system. As far as I can remember, the only Concourse Computer Center (CCC) hacker who was actually a member was myself, the founder, because the hardest single thing was finding space for the LOGO Lab's PDP-11/45 which became surplus around the beginning of 1980 when the Lab was running out of everything. At the time legendary EE professor Jerome Letvin was the head of Concourse, and he was able to find surplus space in Building 20 for it. Any Concourse student was of course encouraged to use it or any of the subsequent systems we scored, computing time was a very scarce resource back then, but its connection to the Concourse program was otherwise very tenuous.
-- after it walked across the street to Bldg. 20, the KA still ran, but had no memory, since its latest memory incarnation was all modified LispM Mem boards, which were given to needy Lispms when the Lab flushed the machine. i think the Concourse hackers have since had access to more of these Mem boards, but they have not (alas!) managed to bring the KA up in full-fledged fashion.
That matches my memory although I wasn't involved in this effort, my impression is that it was treated as an historical artifact to study by the more EE inclined CCC hackers. I also noticed large piles of Knight TV bitmapped monitors and keyboards.
the AIKA went to the concourse computer club folks (an MIT student org.). we told them to tell us if they later decided they didn't want it any more. when peter lothberg came over here to pick up MC, his shipping container had plenty of room left, so we checked to see whether the CCC folks still had it/still wanted it/it still worked in any fashion. we were horrified to discover that Gill Pratt had just sold it to a scrap dealer for its materials, apparently to make room for some sort of solar-car project he was then involved in, without any notice to us whatsoever.
Let's just say that no one who knew him well was surprised when the School of Engineering took the very rare action of vetoing the EECS Department's decision to grant Gill tenure.
Hi PDL!
On 10/11/2018 05:22 AM, dlebling618 wrote:
I don't recall the Morse Code project ever having an 11. It had a fancy digitally controllable radio, but that was on the 2nd floor.
There was a minicomputer in that second floor lab with the Collins radio. It was a small PDP-11 but I can't remember which model. It was about the size of an early desktop PC. Not very many people knew about it or worked with it since it was not a general-purpose machine but really a dedicated part of the hardware for that single project.
IIRC, Ed Black (EHB@MIT-DM) did most of the software work on the PDP-11. The radio speaker output was cabled to some custom hardware that Ed and I built, which interfaced to the -11. The -11 could control the radio to "turn the dials and switches" by software command. The radio was a military-grade Collins product which was typically used in government installations where there was no human present - e.g., remote listening posts in the Arctic, with access via a telephone line and modems. It was the only radio we could find at the time which could be controlled via RS232 and thus by a computer program running in the PDP-11.
There were A-to-D converters to turn the audio into bits, software to do FFTs, PLLs and buffers, etc., etc. The result from processing the audio was a file with a bunch of numbers specifying things like the length of marks (signal present) and spaces (no signal) of the dots and dashes of Morse code. That file was further processed by the DM PDP-10 part of the system written in Muddle. I don't recall exactly how the file got from the PDP-11 to the PDP-10; probably over a standard TTY line. It was all designed to be able to operate in realtime -- but only if the PDP-10 was also dedicated. IIRC, it came very close but didn't quite meet the realtime goal since computers weren't quite fast enough then.
There's a longer discussion of this in:
http://3kitty.org/artifacts/MorseProjectPaper.pdf
One of the diagrams shows how the "PDP-11" on the second floor fit in.
/Jack
I conflated the radio and the PDP-11 into one device.
I'm fairly sure that we moved the processed audio file to the ITS PDP-10 using tftp (or a precursor) via a TTY line.
Yep, I can't think of any other way we might have moved files. Metcalfe hadn't invented Ethernet yet. Extending the PDP-10 or PDP-11 bus from the 9th floor to the 2nd floor would have been ridiculous....especially after the unfortunate incident with the CIA motivated building management to take back my key to the wiring closets......
I've often wondered how technical progress is influenced by such unlikely issues as allocation of office and lab space. If we hadn't been forced to be 7 floors away from our computers, we never would have felt the need to develop such 'client/server' techniques for using systems with small computers interacting with larger ones.
Necessity truly is the mother of invention.
In January 1987, Donald Knuth entertained his CS304 class with a trivia hunt, which included among other assorted Stanfordiana the following question:
Q. List all computers at Stanford that run the TOPS-20 operating system, and tell in what building they are located. 4 points for each correct name and 6 points for each correct location!
A. Various
host
commands identify them as:score
,sushi
, andtruffle
in Margaret Jacks Hall;lear
,othello
,hamlet
andmacbeth
(akalots-a
thrulots-d
) in CERAS;how
andwhy
in the Business School;sumex
andtiny
in the Medical School;sierra
in Durand;turing
(akacsli
) in Pine Hall. There's alsopanda
at Mark Crispin's home; this may qualify as "at Stanford" in some loose sense.
He repeated this in January 1989, with mention of a TENEX machine:
Q. What Stanford computer has its name displayed in stained glass? 15 points!
A. The SUMEX-AIM computer in Stanford Medical School.
@billstclair, you wrote on IRC:
I never programmed in MDL, but I remember PDL & his friend talking about how it learned to generate good machine code for your MDL code. I was writing PDP-11 Fortran to collect morse code data for their intelligent MDL code that parsed it That system ran in ITS on PDP-10s with Imlac graphic terminals for UI
In this GitHub issue you will find some comments about the Dynamic Modeling Morse code project, and the PDP-11. Do you have anything to add?
@billstclair, you wrote on IRC:
I never programmed in MDL, but I remember PDL & his friend talking about how it learned to generate good machine code for your MDL code. I was writing PDP-11 Fortran to collect morse code data for their intelligent MDL code that parsed it That system ran in ITS on PDP-10s with Imlac graphic terminals for UI
In this GitHub issue you will find some comments about the Dynamic Modeling Morse code project, and the PDP-11. Do you have anything to add?
Nothing specifically about the Morse Code project, but I'll relate one thing I remember about the Dynamic Modeling (DM) Lab. Use it if you want.
There was a multi-user maze game that ran on the DM PDP-10s under ITS, using Imlac terminals for display. It used very simple line drawings for maze and players, and allowed you to chase each other around the maze and shoot at each other. It was the earliest first person shooter I'm aware of. I knew only one guy who knew how to start it up, and we always played late at night.
I have reproduced the maze display, as I remember it, but not yet the players or missiles, at jsmaze.com.
Thanks @billstclair.
We have Maze War assembled from source code, both the Imlac part and the PDP-10 part:
https://github.com/PDP-10/its/blob/master/src/imsrc/maze.3
https://github.com/PDP-10/its/blob/master/src/klh/mazser.141
It doesn't run quite well in the Imlac emulator yet, but we're working on it. I believe this software has been used to run the game:
https://github.com/hpalmer/imlac
In January 1987, Donald Knuth entertained his CS304 class with [a trivia hunt] (https://www.saildart.org/TRIVIA.TEX%5B304,DEK%5D), which included among other assorted Stanfordiana the following question:
Q. List all computers at Stanford that run the TOPS-20 operating system, and tell in what building they are located. 4 points for each correct name and 6 points for each correct location!
A. Various
host
commands identify them as:score
,sushi
, andtruffle
in Margaret Jacks Hall;
SCORE was the original Computer Science Department TOPS-20 system. Sushi came along c. 1986; I don't remember Truffle at all.
SCORE's name, like that of MIT-XX, reflected the DEC designation of the system. SCORE was the official Stanford repository for TOPS-20 sources--changes made at any other site on compass were reflected back to SCORE.
lear
,othello
,hamlet
andmacbeth
(akalots-a
thrulots-d
) in CERAS;
These were my babies at LOTS (Low Overhead Timesharing System), which started in 1976 with a single DECSYSTEM-2040, named "LOTS". When a second machine was added a few years later, it was of course names "LESS". By the time I interviewed at Stanford in 1984, they were moving the 2 systems into the new computer room, built from a TV studio on the same level of CERAS, and adding a third system. With the new system came new names: LOTS-A, LOTS-B, and LOTS-C. When we acquired the first customer ship Systems Concepts SC30-M in 1986 (at first called "EPIC"), a contest was held among all the users of the systems (to avoid naming the new machine "LOTS-D") to choose new names. Eventually, the slate of candidate names proposed as "The Rulers" by the student staff of LOTS won: LOTS-A => Lear LOTS-B => Othello LOTS-C => Hamlet Epic => Macbeth Imagine my chagrin when they were introduced in the Stanford Daily as "The Tragedies"...
how
andwhy
in the Business School;
These were managed by Sandy Lerner, co-founder of Cisco Systems.
sumex
andtiny
in the Medical School;
I don't remember the name "tiny". The 2 systems were "SUMEX2065" and "SUMEX2020". The ARPANET/Internet Macintosh developers' mailing list INFO-MAC was hosted on the 2065; for a couple of years I was the moderator.
sierra
in Durand;
SIERRA was the Electrical Engineering department's DEC-20. There was eventually a second, but I can't remember its name to save my life.
turing
(akacsli
) in Pine Hall.
I don't remember it being called Turing, only CSLI (which stood for Center for the Study of Language and Information). Terry Winograd was there, as was one of my former students (Peter "Pierre" King) who took the system manager job.
There's also
panda
at Mark Crispin's home; this may qualify as "at Stanford" in some loose sense.
Mark maintained the hosts.txt file for the campus at SCORE. No one begrudged him the IP address for his (first) 2020.
He repeated this in January 1989, with mention of a TENEX machine:
That would have been the DEC-1077 (i.e., a dual KI-10) at IMSSS, the Institute for Mathematical Studies in the Social Sciences, under the direction of Patrick Suppes.
Umm, so which of these Stanford machines ran ITS?
Nary a one, of course. But SAIL, aka SU-AI, ran WAITS, whose naming was influenced in part by ITS. (I have that on the authority of my friend and former boss Ralph Gorin, who was one of the people who wrote WAITS.)
Maybe Moon is referring to the title of this issue: "Listing all known ITS machines". So the subject of Stanford machines seem out of topic. He'd be technically correct, of course. The best kind of correct. But like most other discussions on the net, we frequently stray from the original subject. I think we can err on the side of inclusiveness when it comes to PDP-10 machines.
Also, I'm to blame myself, because I added some lists of TOPS-20 and WAITS machines to the very first comment.
About concurrent users. I think this is PEEK displaying on a Knight TV, so it would hint at the number of jobs on the AI KA10. (Sorry, there is nothing higher resolution. I asked MIT, but they wanted money to make a better version.)
Source: https://www.csail.mit.edu/node/6625, 9:30 in.
Re your blurry peek display.
Max jobs in ITS 63, including the system job and the core job (but not the null job, which doesn't get its own USR number). Max human users on MIT-AI was usually around 10 or 12 before the system was too bogged down, on days when all 512 KB of core was working.
--Dave Moon
Photos of nearly all MIT ITS machines.
AI PDP-6.
DM KA10.
ML KA10.
.
MC KL10.
AI KS10, the one to the right. Unfortunately not its natural habitat at MIT.
Plus various equipment like Fabritek memory, Stanford keyboard, Knight TV console, GT40, Datapoint, GE Bag-Biter. See #425.
Hello,
Does anyone recognize the machine in this picture? I asked Horn, but he doesn't remember. I think it looks like a KA10 console behind the teletype, but then there's a display where the paper tape reader would usually go.
TK said it's the AI PDP-10. Details here: https://github.com/PDP-10/its/issues/425#issuecomment-469984923
When did the TV-11 go away? Some time between these two messages:
Date: 30 August 1982 05:48-EDT
From: Christopher C. Stacy <CSTACY at MIT-AI>
To: BUG-ITS at MIT-AI
Sometimes when I am using a Knight TV, my screen
gets garbled. I dont know what part of the TV
system (10,11, or 10/11) is losing.
Date: 24 September 1982 09:49-EDT
From: Christopher C. Stacy <CSTACY at MIT-MC>
Subject: ai namedragon
To: CENT at MIT-AI
cc: BUG-ITS at MIT-MC
There being no TV-11 anymore, I retired the name dragon.
Also, since the XGP vanished, this suggests the 10-11 interface was in trouble at this point:
Date: 21 September 1982 18:16-EDT (Tuesday)
Sender: GUMBY at MIT-OZ
From: David Vinayak Wallace <GUMBY at MIT-MC>
To: bug-its at mc
Subject: Whither XGP?
How does ML talk to the Chaosnet? Does it use a 10/11 interface?
A number of people have been screwed (including everybody in 6.001)
because the XGP vanished. Would it be possible to put the XGP on
either ML or MC, at least until large fixed-width fonts exist for
either the dover or whatever new hardware we get?
This is what I know so far.
At MIT.
Outside MIT.
Other PDP-6'es in Cambridge:
TOPS-20 machines at MIT:
WAITS at Stanford: