Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Antique computer chess program website

Author: Robert Hyatt

Date: 11:53:55 09/24/05

Go up one level in this thread


On September 23, 2005 at 22:43:29, Carey wrote:

>On September 23, 2005 at 20:25:01, Robert Hyatt wrote:
>
>>On September 22, 2005 at 13:38:04, Carey wrote:
>>
>>Cool stuff.  Browsing thru chess 4.6 (COMPASS) and Duchess (IBM ASM) was a trip
>>down "memory lane" for me, having programmed on both of those machines/assembly
>>languages way back when.
>
>I thought you might find it interesting, since you competed against both of
>those.
>
>Most people these days find it interesting for historical reasons, but you had
>first hand experience with those systems and the programs.
>
>You've actually sat across the table from those programs, wondering if it was
>going to find a move that CrayBlitz had missed.
>
>
>You know... one of these days, you really should write down a lot of the side
>stories about those tournaments.  All the interesting bits of trivia and
>encounters and so on.  I think a lot of people today would get a kick out of it.
>
>I've browsed through the CCC and rec.games.chess.computer archives and found old
>messages from you telling about this experience or that encounter, etc. and it
>was a lot of fun to read.
>
>You are the only one still around after all this time.  And not only are you
>still interested, you are still productive and competitive!!  I have *no* idea
>how you've managed to keep your interest in computer chess this long!!
>Incredible.
>
>
>>The stuff from "blitz" seems to be essentially worthless, unfortunately...
>
>Yeah.  I know.
>
>But it's all that he found, and I figured a little bit of 'worthless' was better
>than nothing.
>
>
>>There used to be a guy on ICC that played and had a version of blitz.  I think
>
>Really...?
>
>>the account was something like "dark blue" rather than "deep Blue" or something
>>along those lines.  He had the 1978 version and had moved parts of it into some
>>sort of high-powered PCI graphics card...  I will see if I can figure out who
>
>A graphics card... intriguing.  I've thought of using those things for numerical
>computations, but the limited precision just wasn't enough to do FFTs, etc.
>
>>it
>>was and if he is still around.  He definitely had the complete source.
>
>Oh cool.  That would be cool.
>
>
>I assume it was probably FORTRAN 66, right?  If so, it would probably still
>compile, too.
>
>
>>I could probably scan in a version of Cray Blitz that I still have a printed
>>copy of.  It was probably the 1986 WCCC version, so it is pretty old, and is not
>
>Hey, that would be cool!  It doesn't matter that it's hardwired for the Cray.
>
>I didn't know you still had a printed copy.  (By the way, a couple of authors
>I've talked to are in the same situation... No electronic copies, just some old
>printouts.)
>
>(I don't suppose there is a cray emulator around, so people could have fun...
>Probably not.  Maybe Qemu emulator could be extended for the Cray, but it
>wouldn't be worth the effort.  Still, it would be fun... I mention on the page
>that some cell phones today have more computing power than many of the computers
>from the 70s and 80s...  (Exluding the Crays, of course.)  It would be real
>amusing to get CrayBlitz ported to run under a cell phone... [laugh!]  No, I'm
>not going to do it and I know there are difficulties etc., but it sure is a
>funny mental image.  Condensing a room full of Cray down to a phone...)
>
>


This could actually run, since everything was in FORTRAN, and then selected
functions were rewritten in assembly.  But the FORTRAN was always kept as well
for testing on non-crays.

Will be a royal pain to debug I am sure, because the printout is in a landscape
two pages per 8.5 x 11 page, which means the fonts are a bit small, and I
suspect the OCR software will have a problem.  This stuff works better on text,
where you can look up words in a dictionary to see if your "guess" for a
specific letter makes sense.  For FORTRAN, this won't work.  But I'll see what I
have.  I might just scrap the cray assembly part since the fortran is really
what could be used to make an executable...



>That would definetly be a worthwhile addition to the collection.
>
>
>>that recent since the program played thru 1994 and was modified through that
>>date.  The danger of scanning is that it won't run of course, since things line
>>1 (one) and l (lowercase l) get screwed up with OCR software (to name just a
>>single case, there are lots of others).
>
>
>I know about the scanning experience!  [shudder]  It is not fun at all.
>Actually, I found a lot of the stuff I've scaned went okay...  Especially if you
>let it learn the font.  The hard part was trying to scan old magazine
>photocopies.  Poor quality magazine printing to begin with, followed by a poor
>photocopy at an angle.  Trying to tell ( from [ from {, ; from : and so on...
>Lots of fun.
>
>
>
>For the record, we do have one more significant tournament program comming for
>certain.  It just hasn't arrived yet.  (Don't want to name anybody.)  The
>authors agreed to it, they just haven't gotten around to doing it yet.  Don't
>know where they live.  For all I know they live right on the gulf coast and have
>significantly more important things on their mind.
>
>And a couple of others that might, if they get around to it.
>
>It's all pretty time consuming, since I'm doing this in my spare time.  A few
>others are helping, but I seem to be doing most of the hunting.  It was my idea,
>so I guess I can't really complain....[grin]  And it's not like I have to get
>this done by any certain time.
>
>Due to the weather, I doubt I'll get all that much done over the weekend.  I
>don't even live at the gulf coast, but the local weather guys are still saying
>anywhere from 5 to 15 inches of rain.  One suggested possibly even more,
>depending on the path and circumstances.  Just have to wait and see, I guess.
>
>
>Anyway, yes, if you can find a full copy of Blitz, we would definetly like to
>post it.
>
>And yes, if you can scan in CrayBlitz, we'd like that too.  Depending on the
>quality of your OCR program, it should work out okay and the finished product
>would be good enough.  We wouldn't need the scan image itself.  (I mention that
>because when I did my scanning and OCR'ing, a couple of the ocr demo programs I
>tried did not do too well at all.)



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.