Author: Robert Hyatt
Date: 23:04:02 09/24/05
Go up one level in this thread
On September 24, 2005 at 22:48:29, Carey wrote: >On September 24, 2005 at 14:53:55, Robert Hyatt wrote: > >> >>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. > >That's good to hear. > > >>Will be a royal pain to debug I am sure, because the printout is in a landscape > >Well... > >It's been years since I've done any Fortran, but if you want to save yourself >some time, you could just scan it in and let everybody else work on the >debugging. > >I don't know what your program looks like, so I certainly can't say how easy >it'd be to debug. For all I know, all of your variables were composed of some >combination of '1', 'l', 'O' and '0', in which case it'd be darn near impossible >to debug...[grin] Once someone told me "this is the cleanest-written chess program I have ever seen written in FORTRAN." but then we started optimizing things for the Cray. It still was not _that_ bad after we finished, but some of the code looks very strange for a non-cray programmer... I'll see what I can do. If I recall, it was about 30K lines of code, however, so this is not going to be a quick and dirty deal. The variable names were reasonable, although everything conformed to early FORTRAN standards (there was no recursion, variable names were limited in length, or at least the first 6 or 8 letters (6 I think) had to be unique), etc. More when I get to the office and track this printout down... It dates to somewhere in the 1986-1989 time frame, but where exactly I am not sure I can say. I ran across it a few months/years back, and am pretty sure it is still in a ring binder on one of my bookshelves, or else in a file folder... would also be interesting to see how fast it runs today, compared to speeds on the old Cray hardware... I will add that the code was specifically modified in several places to vectorize, but it ran like a dog on a non-vectorized box since the memory accesses were so much slower than on a cray (still true today of course).
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.