Author: Robert Hyatt
Date: 20:03:52 06/17/00
Go up one level in this thread
On June 17, 2000 at 22:30:18, Christophe Theron wrote: >On June 16, 2000 at 22:32:28, Robert Hyatt wrote: > >>On June 16, 2000 at 20:54:22, Christophe Theron wrote: >> >>>On June 16, 2000 at 20:15:58, Robert Hyatt wrote: >>> >>>>On June 16, 2000 at 15:02:31, Oliver Roese wrote: >>>> >>>>>I recall, i saw a printout of horrible quality in a book, some long time ago. >>>>>So i know the code is available for the public. But it seems not to be found on >>>>>the net somehere. >>>>>Or am i wrong here? >>>>> >>>>>Oliver Roese >>>> >>>> >>>>Note that it was written in Compass (CDC assembly language). I am sure it is >>>>laying around somewhere... I used to have a copy of it myself. However, it was >>>>very hard to read... >>> >>> >>>I remember that I have read somewhere that Chess 4.x was entirely written in >>>Fortran 77. >>> >>>Well you must know better than me, as you have met the authors of the program... >>> >>> >>> Christophe >> >> >>It was definitely written in Compass. You will even find a reference to this >>in Chess Skill in Man and Machine. They wrote a few cute macros to help cut >>the size of the code down, but it was hard to read. > > >OK, so the magazine was definitely wrong... > > > Christophe > > Definitely wrong. As I mentioned elsewhere, the program was written in 1974-1974. _way_ before there was anything known as "FORTRAN 77". By the time there were F77 compilers, chess 4.x was history and chess 5.0/NuChess was written by Slate. It was written in FORTRAN, mainly so he could run it on a Cray as I was doing... That's the main reason for early FORTRAN programs. FORTRAN existed on more computers than any other language. On a Cray, in the late 1970's, you had to choose between (A) FORTRAN (B) FORTRAN (C) CAL (Cray Assembler Language). Those of us wanting to "go fast" went the assembler route once we had a target architecture firmly in mind... wrecks portability of course... but it can be _way_ faster... > >>Dave Cahlander still works at Cray I think (he worked for CDC and wrote the >>Compass assembler way back when). He might be a contact to get a copy of the >>source code. But in any case, I have personally been thru the code, and did >>some porting work when Harry and I were running it on a Cyber 176 at Livermore >>as a tune-up opponent for various chess events. Cray Blitz and most other >>programs were FORTRAN. But Chess 4.x and Duchess were pure assembler.
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.