Author: Dann Corbit
Date: 15:48:11 11/24/03
Go up one level in this thread
On November 24, 2003 at 18:42:35, Peter Skinner wrote: >On November 24, 2003 at 18:37:26, Dann Corbit wrote: > >>On November 24, 2003 at 18:32:43, Peter Skinner wrote: >> >>>On November 24, 2003 at 18:11:49, Dann Corbit wrote: >>> >>>>On November 24, 2003 at 17:55:10, James T. Walker wrote: >>>> >>>>>On November 24, 2003 at 17:42:59, Robert Hyatt wrote: >>>>> >>>>>>I have just released the source for version 19.6. It has two changes >>>>>>of interest: >>>>>> >>>>>>1. The compile problems for windows are gone. It should compile cleanly >>>>>>for SMP and non-SMP versions, as is. >>>>>> >>>>>>2. Eugene has modified the egtb probe code to reduce the decompression >>>>>>indices memory requirement by nearly 50%. For the tables I use, the old >>>>>>version needed 177mb, the new version needs 100mb. The new version is also >>>>>>a couple of percent faster due to another tweak we tried... >>>>> >>>>>To where did you release it? Why are there no more ".exe" versions of Crafty at >>>>>the FTP site? Since I can't compile it myself I'm out in the cold. I hate >>>>>modified versions of a program without having the original to compare it to. >>>> >>>>I'll put up a binary in a few minutes. >>> >>>Hi Dan, >>> >>>When I try to compile _any_ version of Crafty with MSVC++ 6.0 w/sp5, for some >>>reason when it plays I can not get over 95% cpu. Your version has no problems. >>> >>>I use the default makefile, and the commandline nmake wcraftyx -f makefile.nt >>> >>>Is there something I am doing wrong? >> >>Don't know. I use my own build scripts. >> >>>Should I install the processor pack from Intel? I would but I have no clue where >>>to get it anymore. >> >>I had it installed when I was using VC++ 6.0. I use MS VC++.NET 2003 now. > >I was thinking of purchasing the newer version. Do you know if the default >makefile for would work with it without modification? I always have to fiddle with stuff a little bit. Eugene has some sort of magic next-generation compiler at his disposal, since he is a compiler writer at Microsoft. It never compiles for me right out of the box. But the few little problems are always obvious and easy to solve. Some of the problems I generate myself, since I don't compile individual source modules. This is my crafty source startup: #ifdef _WIN32 #include <windows.h> #include <process.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #endif #include "chess.h" #include "lock.h" #include "data.h" #include "epddefs.h" #include "epdglue.h" #include "epd.h" #include "evaluate.h" #include "vcinline.h" #include "x86.c" #include "boolean.c" #include "swap.c" #include "attacks.c" #include "evaluate.c" #include "make.c" #include "unmake.c" #include "movgen.c" #include "quiesce.c" #ifdef DO_EXTENSIONS #include "searche.c" #else #ifdef DO_BM_EXTENSIONS #include "searchb.c" #include "sing.c" #else #include "search.c" #endif /* DO_BM_EXTENSIONS */ #endif /* DO_EXTENSIONS */ #include "next.c" #include "searchr.c" #include "repeat.c" #include "nextr.c" #include "history.c" #include "nexte.c" #include "utility.c" #include "valid.c" #include "searchmp.c" #include "thread.c" #include "enprise.c" #include "book.c" #include "data.c" #include "drawn.c" #include "edit.c" #include "epd.c" #include "epdglue.c" #include "init.c" #include "input.c" #include "interupt.c" #include "iterate.c" #include "main.c" #include "output.c" #include "phase.c" #include "ponder.c" #include "preeval.c" #include "resign.c" #include "root.c" #include "learn.c" #include "setboard.c" #include "test.c" #include "time.c" #include "validate.c" #include "annotate.c" #include "option.c" #include "analyze.c" #include "evtest.c" #include "bench.c" #include "testepd.c" #ifdef DO_BM_EXTENSIONS #include "hashb.c" #else #include "hash.c" #endif /* DO_BM_EXTENSIONS */ #include "probe.c"
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.