• Please review our updated Terms and Rules here

Pdp8 vs pdp11 chess match is GO!

R

Roe

Experienced Member
Joined
Dec 28, 2012
Messages
240
Location
Regina, Canada
At least from the pdp11 side. My 11/83 is up and running, connected to the net, and is ready for the game.

Instead of posting the moves to the forum for the audience, I've set up a server for a (hopefully) interested audience. It is available for testing now (hint) and should display a welcome message.

Telnet to 24.72.2.225 port 8000 and let me know if you have problems.

So, let's figure out the when. I don't think there are any major football games this weekend, and I'm available both Saturday and Sunday.

Comments or suggestions (especially from the Pdp8) are very welcome.
 
This reminds me of the movie Computer Chess.

Personally I didn't enjoy the movie, but critics seemed to love it.

Lots of vintage hardware in the movie, including a PDP-11.

It's streaming on Netflix and Hulu.
 
I gave it a try:

Code: 
[smpUnibodyMBPearly2011:~] spereira% telnet 24.72.2.225
Trying 24.72.2.225...
Connected to static24-72-2-225.r.rev.accesscomm.ca.
Escape character is '^]'.


2.11 BSD UNIX (rap83.rapnet.com)

login:

That looks promising. How will we log in to watch?

smp
 
smp said:
I gave it a try:

Code: 
[smpUnibodyMBPearly2011:~] spereira% telnet 24.72.2.225
Trying 24.72.2.225...
Connected to static24-72-2-225.r.rev.accesscomm.ca.
Escape character is '^]'.


2.11 BSD UNIX (rap83.rapnet.com)

login:

That looks promising. How will we log in to watch?

smp
Click to expand...

I think you'll want to include the port after that.

Code: 
$ telnet 24.72.2.225 8000
Trying 24.72.2.225...
Connected to 24.72.2.225.
Escape character is '^]'.

Welcome to the vintage computer chess match!

Here we pit a pdp8 against a pdp11.

Stay tuned for scheduling announcements.

Kyle
 
I used putty. And got to the "Stay tuned for scheduling announcements".
 
Stupid, stupid Windoze and MS-DOS. I can telnet in alright, but MS-DOS doesn't know how to LF. Oh well, my wife's computer should be online by now. I'll use that.
 
antiquekid3 said:
I think you'll want to include the port after that.

Code: 
$ telnet 24.72.2.225 8000
Trying 24.72.2.225...
Connected to 24.72.2.225.
Escape character is '^]'.

Welcome to the vintage computer chess match!

Here we pit a pdp8 against a pdp11.

Stay tuned for scheduling announcements.

Kyle
Click to expand...

Got it. Thanks. That worked fine for me.

smp
 
So far, this is all very encouraging. For the techies among us (100%?) the service on port 8000 runs a tail -f command on /tmp/chess.out. The game is going to run with:

chess | tee -a /tmp/chess.out

I'm going to make sure that the board is printed after each move.
 
Roe said:
At least from the pdp11 side. My 11/83 is up and running, connected to the net, and is ready for the game.

Instead of posting the moves to the forum for the audience, I've set up a server for a (hopefully) interested audience. It is available for testing now (hint) and should display a welcome message.

Telnet to 24.72.2.225 port 8000 and let me know if you have problems.

So, let's figure out the when. I don't think there are any major football games this weekend, and I'm available both Saturday and Sunday.

Comments or suggestions (especially from the Pdp8) are very welcome.
Click to expand...

My 8m is ready to go, but this weekend is NOT good :( (Mother's Day). How's the following weekend look?

Also, I'm having trouble connecting; I get a "Could not open connection to host, on port: 8000:Connect failed"
Finally, I'm a little sketchy on the details of how I will upload my machine's moves.
 
The weekend of the 17th works for me. Saturday afternoon good with everyone?

You probably tried connected after I shut down the system. I am convinced everything works correctly, as several successful connections came in.

As for uploading your machines moves, it's probably simplest if you just email them to me, and I'll enter them into the pdp11.

I'm looking forward to the game!
 
It might be interesting to include Northwestern Chess in the competition. My exposure was with Chess 3.0 back in the 70s. It was quite a CPU hog--and it used one side of the operator's console to boot. Still, endless hours were spent on the machine floor at CDC playing the thing. Today, you'd probably have to use the emulator.

And yes, programming language fans, it was coded in FORTRAN. :)

Even though the 6600 was a 10MHz machine, I suspect it would beat the pants off of a PDP-11 or 8...
 
I'll be the first to confess that memories of what I did decades ago may be subject to bitrot. I did recode parts of the Chess 3.0 code to work with the ZODIAC operating system (a derivative of TCM, which, in turn was based roughly on SCOPE 3.1.6) back then--but some system calls were different. I do remember a PP display driver called "CHD" that was part of the package also, so you had to EDITLIB the thing into your system. I don't see any mention of that in the 4.6 code.

It could have all been COMPASS, it represented so little of the stuff that I was doing, that it didn't make a big impression at the time. To make matters more confusing, you could mix COMPASS and FORTRAN source decks and FTN would recognize the difference and invoke COMPASS if required.

A little later, I ported the code for Adventure/Colossal Cave, from a DECSystem 10 tape that was given to me by a friend who was a DEC CE. That was more of a hit in the ranks and was pretty much straight FORTRAN with DEC extensions--I recall a 9-track tape with 5 ASCII characters packed per 36-bit word.

When you've been allocated 4 hours of dedicated machine time at 2 AM, you'll do all sorts of things to relieve the boredom of watching your code compile.
 
Chuck(G) - thanks. Since posting that reply I've been reading and searching. Here's what I've found...

Another article confirms v3.6 seems to indeed have been a combination of FORTRAN and assembly. Perhaps this is the one you recall? v4.6 was a complete re-write - in COMPASS.

Some time later, in 1978 a PASCAL version (dubbed 0.5) was published by the same author (Larry Atkin and Peter W. Frey). A thread marking it's internet debut is here and the source code is available here. Since that source is OCR'd and and written in an ancient dialect, it may take considerable adaptation run nowadays. One would think that because of it's heritage it would represent qualitative progress in performance over the predecessors, but that's surmise.

More... Others have been working on the 0.5 version. There may yet be working PASCAL source code out there, but links to it in dropbox are dead so far.
 
Last edited:
Roe said:
The weekend of the 17th works for me. Saturday afternoon good with everyone?

You probably tried connected after I shut down the system. I am convinced everything works correctly, as several successful connections came in.

As for uploading your machines moves, it's probably simplest if you just email them to me, and I'll enter them into the pdp11.

I'm looking forward to the game!
Click to expand...


The 17th sounds good to me! I'm in California, so we need to define 'afternoon'; 11:00 am PDT works for me, but I'm open to other times. Just let me know.

My 8m is chomping at the BIT (hardee, har, har :D)
 
RSX11M+ said:
Chuck(G) - thanks. Since posting that reply I've been reading and searching. Here's what I've found...
Click to expand...

I thought I remembered that one right--but after 40-odd years, the mind plays tricks. :) Dewey did beat Truman, right? :lol:

The mixed FORTRAN-COMPASS meme was very common in those days. Even the FORTRAN compilers used it.

FTN on OPT=x optimization could do some very good instruction scheduling, depending on your nerve. For example, on the right loop, it could code prefetch of the next iteration's operands at the bottom of the loop, thereby overlapping the fetch with the branch time. There were also some cute bit-twiddling tricks that it knew.

Easier and shorter to let a program manage the scheduling than doing it all by hand--and programmers are inherently lazy. :)

RUN was the older compiler that had the charm of being able to stack a FORTRAN/COMPASS mix behind a single JCL statement that said merely "RUN."
 
AteEm said:
The 17th sounds good to me! I'm in California, so we need to define 'afternoon'; 11:00 am PDT works for me, but I'm open to other times. Just let me know.

My 8m is chomping at the BIT (hardee, har, har :D)
Click to expand...
Which bit?
 
You must log in or register to reply here.
Back
Top