• Please review our updated Terms and Rules here

My 8m passes the Chess.bin test....I didn't!

Just for the record, I think that a 4 bit machine could probably beat me at chess. A PDP-8/M probably could also beat me at kickboxing. :)
 
OK, I'll run the 11/73 without lobotomization, the pdp8 will play white. Does Wednesday afternoon about 13:00 CST sound like a reasonable start time?
 
Roe said:
OK, I'll run the 11/73 without lobotomization, the pdp8 will play white. Does Wednesday afternoon about 13:00 CST sound like a reasonable start time?
Click to expand...

Sadly, I drive a desk from 8:00 to 5:00 (PST) in order to pay for my DEC habit, er, hobby:eek:
So, this will need to be a late evening event for those farther east...or, it will need to be a weekend matchup.
Either way is good for me.

Mike
 
Can we get a technical rundown of the two competitors? :) (CPU, speed, RAM, anything of interest).
 
Yeah, now that I think about it, the weekend is everyone's best choice. How does Saturday afternoon sound?
 
I'm going to put my pdp11/83 into the ring. It's a KDJ11 CPU at 18Mhz with 4MB ECC PMI memory in a BA123 "airplane trolley" box, with a CMD CQD220/TM scsi controller. It will be running V7 unix with the standard chess game from that distribution. Short of the 11/93, this is the fastest 11/xx ever made by DEC.

Mike, once I found out that the pdp8 was coming up with a move every couple of minutes, I just had to go with my best box. It's still probably going to take 10-15 minutes for some moves -- as I said, V7 chess is NOT known for speed.
 
barythrin said:
Can we get a technical rundown of the two competitors? :) (CPU, speed, RAM, anything of interest).
Click to expand...

Pretty basic 8m: 4 board CPU, 8K core (3 board set, nearly 2 pounds, REAL donut core, no pansy CMOS stuff here), and serial interface. Full front panel, though, with plenty of blinkin' lights and switches..oooh, la, la.
Instruction cycle time is 1.2 or 1.4 microsecond depending upon instruction. So, I guess it's about a 0.8 MIPS machine :)
Using my ASUS netbook as a dumbterm hooked to the serial card in the 8m through a RadioShack USB/Serial converter.
The chess program is all of 4K (no wussy C++ megaprogram for my machine) that I downloaded from bitsavers.
 
Roe said:
I'm going to put my pdp11/83 into the ring. It's a KDJ11 CPU at 18Mhz with 4MB ECC PMI memory in a BA123 "airplane trolley" box, with a CMD CQD220/TM scsi controller. It will be running V7 unix with the standard chess game from that distribution. Short of the 11/93, this is the fastest 11/xx ever made by DEC.

Mike, once I found out that the pdp8 was coming up with a move every couple of minutes, I just had to go with my best box. It's still probably going to take 10-15 minutes for some moves -- as I said, V7 chess is NOT known for speed.
Click to expand...

Ruh, oh. Why do I think I'm going to be feeling a lot like Peyton Manning after this is all over...
 
AteEm said:
Ruh, oh. Why do I think I'm going to be feeling a lot like Peyton Manning after this is all over...
Click to expand...
Disabling CACHE will help with that machine too... though not as much as with the 73. :)
 
AteEm said:
Ruh, oh. Why do I think I'm going to be feeling a lot like Peyton Manning after this is all over...
Click to expand...

LOL for the Ruh,oh!

The 11/83 will play the same game as any other pdp11, just won't be quite as boring. Looking forward to Saturday afternoon, it should be interesting. I think it's a bit of a miracle that someone managed to squeeze _any_ kind of chess game into 4K of memory.
 
Roe said:
LOL for the Ruh,oh!

The 11/83 will play the same game as any other pdp11, just won't be quite as boring. Looking forward to Saturday afternoon, it should be interesting. I think it's a bit of a miracle that someone managed to squeeze _any_ kind of chess game into 4K of memory.
Click to expand...
I agree on that. It is amazing they were able to cram any type of chess game (and this one is decent) into 4k.
 
Roe said:
Saturday at 14:00 CST (12:00 PST) sounds perfect.
Click to expand...

Uh, Houston, we have a problem.

In preparation for Saturday's upcoming battle, I was 'sparring' with the program last night and well into the game when the machine took my rook by sliding right through my knight which was in the path to the rook!! I reset all the pieces and carefully re-played all the moves (all the moves were still displayed on my netbook) to make certain I had not mis-positioned any pieces; I found no errors.

I'm going to reload the program with the BIN loader tonight and attempt the same set of moves. If this error repeats I think we done before we've started.

In my web searches I had noticed instructions for a version called CHEKMO II, but never found the binaries. If someone could point me to those, that might be a backup if the program I have proves to be flawed.

Mike
 
Pshaw! How about a real cross-platform world-wide chess tournament; hook 'em up to the Internet and let them play by themselves, against various opponents...

I could probably fire up one of my ancient Cromemcos... ;-)
 
Okay.
All is well after reloading the program.
Played the same game and the 8m made a DIFFERENT move when it got to the same place in the game.

I will be posting White's first move on Saturday at High Noon (well, PST, anyway:))
 
I'm curious - do you know the program to be completely deterministic - or is there a random element?
 
RSX11M+ said:
I'm curious - do you know the program to be completely deterministic - or is there a random element?
Click to expand...
I've been curious about that, too.
With only two games played, and only parts of those games repeated, it APPEARS to be deterministic.
But, I think it's premature to say for certain.
 
Roe, is there time for me to setup a shadow system?

I can't run a PDP-8, but I might have time to load up a PDP-11 with unix v7 to follow along... (not to play)

I'd like to know as many particulars as you think it might take to mimic your load as precisely as possible. So far, the images of v7 I've tried won't execute on hardware much newer than an 11/45 without error.
 
Last edited:
You must log in or register to reply here.
Back
Top