I figured out my own problem, for those having the same issue; you can't compile your code with cygwin gcc and run with winboard even if the program has access to the cygwin.dll; you can however compile with gcc-3 -mno-cygwin in cygwin and run your program with winboard. Im also finding that ...

Hi all Ive reached the point in writing my chess program that its time to connect to winboard. Im getting an error of the sort "second chess program (myprogram) exited unexpectedly". I dont know how to have winboard save the errors; has anyone out there written an engine and faced this same error ...

Thanks for the responses ill try them out. The "usefulness" stems from me wanting to try some stochastic monte carlo stuff even though it has not been shown to work well for chess.

thanks Geko, thats a fast solution; unfortunatly I was hoping this function would work for 64bit numbers and it would require too much memory for tables that big. may look into the rotation idea.

Im looking for a bitposition hack that does the following quickly

suppose I have a bitfield 0b011010

Im looking for a quick function to return the position of one of the ones randomly ie

randomsetposition ( 0b011010 ) { // return 1, 3, or 4 randomly

theres got to be a better way than looping or ...

quick questions about q-search. I have been studying the code of TSCP and winglet and there is part of the q-searching algorithm I cant get my head around. So qsearch is called when you reach a specified depth in alphabeta; the goal is of course to look for a quiet position because if you're in the ...

anyone else having trouble accessing http://www.sluijten.com/winglet/

such a great site, hope it only down for maintenance

So you do cut off branches that fail low. I am glad I asked this question because several responses that I have gotted (and am grateful for) do not agree. User923005 says that PVS is "not about getting more cutoffs". So which is it, both seem like sensible ways to get a preformancs increase.

ahh, so we save time and effort be not having to replace alpha hundreds of thousands of times. but you should search the same number of nodes as alphabeta, potentially more if you get fail lows. Im a bit suprised you can get up to 10% speedup with this detail but I dont have intuition about how ...

ok I think I am starting to wrap my head around this PVS after reading these responses and also rereading the Moreland page. like regular alphabeta with PVS you cut when evaluations are above beta, but now you also cut if evaluations are less than a certian value. Now that I think I understand it a ...