TalkChess.com

Hi Martin, can you please provide this build for download or send me a version? Maybe it is something compiler relevant? thanks Werner Hi Werner, I certainly could but I made some changes and I'm not sure the fix is ok. Also my compiles are slower than Daniel's, so I would prefer to wait for what V...

Werner wrote:Hi Vladimir,
why does 64bit crash and 32bit not?

Sorry, I develop and test on 32-bit only, can't even reproduce this bug at the moment.

Updated to 13.0.

supersharp77 wrote: GreKo 12.5. 0.0
Crafty 20.13 wb. 4.0

Could you please use the latest version of GreKo?
Version 12.5 is obsolete.
Current version is GreKo 12.9.

Joost Buijs wrote:Node-count doesn't mean anything with respect to playing strength.

Yes, remember early Rybkas with their low nodecount.
More to say, depth of search also means almost nothing. Remember Junior

matthewlai wrote:I happen to have quite a few engines in my gauntlet, so I ran a search from start position on all of them. All single threaded.

This approach can be a little bit misleading if different engines count nodes in different ways.

Thank you for information, please send me the games. Anyway, self-play can give only a rough estimation of improvements in strength, I know.

Which time control did you use?
At time controls besides "N moves in M minutes" Greko 12.9 is similar to 12.8a.

No, this is not an error. The compiler does not understand that negative "to" variable will never be used as an index in array because of while(x) condition. #define TRACE_DIR(dir, Shift, delta) \ x = Shift(m_single[from]); \ y = 0; \ to = from + (delta); \ while (x) \ &#...

Checked gcc version again. It defaulted to 4.6 even though 4.9 was installed. I set the default gcc and g++ version to 4.9 and voilà... a working GreKo 12.9. :D It's no longer playing silly moves now. Got some warnings about bitboards.cpp, though: g++ -DNDEBUG -Wall -O3 -c -o bitboards.o bitboards....