Moderator: Ras
I may make the changes for winboard compatibility, but at present it does not ponder since threads do not work the same way they did on the Cray... This is the 1991 version we used to play in the ACM event of that year, although I am not sure this is the "final version" so there might be a bug or two or three lurking...swami wrote:Not sure If you visit the programming forum, but Bob has just released the source code for Cray Blitz, and Jim Ablett made a compiled version.
http://64.68.157.89/forum/viewtopic.php ... =&start=40
It's only a console - no Winboard support, Are there any Fortran programmers who is willing to implement the added protocol into it? It seems many people have forgotten the language, found it hard to go through or have no interest in it...
So over to Fernando, Tony and Thorsten, they could be key motivators in helping or finding out the programmers...
Thanks, Bob. I think many people would be very much interested in WB Cray Blitz no matter whatever version it is, as long as it plays chessbob wrote:I may make the changes for winboard compatibility, but at present it does not ponder since threads do not work the same way they did on the Cray... This is the 1991 version we used to play in the ACM event of that year, although I am not sure this is the "final version" so there might be a bug or two or three lurking...swami wrote:Not sure If you visit the programming forum, but Bob has just released the source code for Cray Blitz, and Jim Ablett made a compiled version.
http://64.68.157.89/forum/viewtopic.php ... =&start=40
It's only a console - no Winboard support, Are there any Fortran programmers who is willing to implement the added protocol into it? It seems many people have forgotten the language, found it hard to go through or have no interest in it...
So over to Fernando, Tony and Thorsten, they could be key motivators in helping or finding out the programmers...
It does have a working book and search, although it is not so strong on a PC because it was really designed around vectors. The assembly language might be interesting to those interested in machines like the Crays...
That's the reason Carey was interested in resurrecting this thing. All the old source files were lost years ago (around 1995). I eventually found one old listing of CB, and a _really_ old listing of blitz 6.9 (circa 1978-1979 somewhere). I sent those to Carey who wanted to try to scan/OCR them. It was pretty hopeless. We worked on it off and on but the OCR software is not anywhere near good enough to tackle this. Luckily someone found an old copy on a backup up at the alabama supercomputer center. So at least we had a working 1991 version of Cray Blitz. Took some work to make it compile cleanly but it finally happened. was interesting to read over the code to see what I had forgotten. For example, I had forgotten we had played around with null move R=2, although we used "1" by default. I had forgotten we had switched to recursive null-move (more than one null in a path, just never two in a row). It was interesting to read back thru the stuff... sort of like a trip down memory lane...swami wrote:Thanks, Bob. I think many people would be very much interested in WB Cray Blitz no matter whatever version it is, as long as it plays chessbob wrote:I may make the changes for winboard compatibility, but at present it does not ponder since threads do not work the same way they did on the Cray... This is the 1991 version we used to play in the ACM event of that year, although I am not sure this is the "final version" so there might be a bug or two or three lurking...swami wrote:Not sure If you visit the programming forum, but Bob has just released the source code for Cray Blitz, and Jim Ablett made a compiled version.
http://64.68.157.89/forum/viewtopic.php ... =&start=40
It's only a console - no Winboard support, Are there any Fortran programmers who is willing to implement the added protocol into it? It seems many people have forgotten the language, found it hard to go through or have no interest in it...
So over to Fernando, Tony and Thorsten, they could be key motivators in helping or finding out the programmers...
It does have a working book and search, although it is not so strong on a PC because it was really designed around vectors. The assembly language might be interesting to those interested in machines like the Crays...
Strength do not matter, there are hundreds of engines that play at 1400's, I'd guess that there are engines out there that don't ponder. It's just that people would be interested in it for sentimental reason, or have fascination for collecting older programs, more so knowing Cray had been a tournament winner once (world champs?), would be delighted if it comes with added protocol support.
Yeah... I pestered you about it for close to 10 years, I guess.... The first time I asked you were pretty sure you had a copy on a tape cartridge but had no way to read it. Later you said the tape wasn't readable.bob wrote:That's the reason Carey was interested in resurrecting this thing. All the old source files were lost years ago (around 1995). I eventually found one old listing of CB, and a _really_ old listing of blitz 6.9 (circa 1978-1979 somewhere).
Yeah, OCRing Fortran is pretty tough. Especially when you throw in the automatic variable declarations.I sent those to Carey who wanted to try to scan/OCR them. It was pretty hopeless. We worked on it off and on but the OCR software is not anywhere near good enough to tackle this.
) I learned my lesson on OCRing Fortran!!) I'll probably follow up on that eventually, just to see if he happened to find an electrnoic copy.Luckily someone found an old copy on a backup up at the alabama supercomputer center. So at least we had a working 1991 version of Cray Blitz. Took some work to make it compile cleanly but it finally happened. was interesting to read over the code to see what I had forgotten. For example, I had forgotten we had played around with null move R=2, although we used "1" by default. I had forgotten we had switched to recursive null-move (more than one null in a path, just never two in a row). It was interesting to read back thru the stuff... sort of like a trip down memory lane...