Search found 17544 matches

by hgm
Sat Apr 20, 2019 7:34 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: No standard specification for Perft
Replies: 22
Views: 417

Re: No standard specification for Perft

Well, you must be doing it wrong, if you use the same move generator for a different purpose...
by hgm
Sat Apr 20, 2019 5:59 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: No standard specification for Perft
Replies: 22
Views: 417

Re: No standard specification for Perft

Exactly. If the engine uses a completely different move generator in QS, a perft of the full-width search would not test it at all. For those who do want to use perft for comparing efficency of various move-generation algorithms and board representations: note that a real engine usually does not mak...
by hgm
Fri Apr 19, 2019 8:47 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: No standard specification for Perft
Replies: 22
Views: 417

Re: No standard specification for Perft

If you want perft to be a representative measure of the speed of an engine it should indeed focus more on capture-only generation. Perthaps it would be useful to define a function perft(N, M) which does N ply of full-width search followed by M ply of captures only. Most people use perft only for tes...
by hgm
Thu Apr 18, 2019 6:25 pm
Forum: Computer Chess Club: General Topics
Topic: Xilinx UltraScale Questions
Replies: 4
Views: 194

Re: Xilinx UltraScale Questions

It would be interesting to know how powerful a TPU one could make from these things, for simulating neural nets.
by hgm
Thu Apr 18, 2019 2:06 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Efficient capture generation in the game of Thud
Replies: 2
Views: 171

Re: Efficient capture generation in the game of Thud

With a mailbox board a 'neighbor table' could be useful: this would, for each occupied square, contain the number of the nearest occupied square in each of the eight directions. It is not very expensive to update such a table incrementally, if you postpone the update to the point where you are sure ...
by hgm
Thu Apr 18, 2019 4:15 am
Forum: Computer Chess Club: General Topics
Topic: The problem with adjudication based on score
Replies: 26
Views: 1246

Re: The problem with adjudication based on score

The simplest solution is to never let the engine print very negative scores. Always divide the score by 10 when it is negative before printing. This should gain you some Elo.
by hgm
Wed Apr 17, 2019 5:30 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Winboard state machine diagram
Replies: 3
Views: 381

Re: Winboard state machine diagram

The command groups are usually work-arounds for bugs in some historic engines. (Which I consider a very bad idea.) But they were designed in such a way that they would do what was desired if all the commands in it would individually do what the specs say they should do. It is just that they also wor...
by hgm
Tue Apr 16, 2019 4:14 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Winboard state machine diagram
Replies: 3
Views: 381

Re: Winboard state machine diagram

I have always felt that the state diagram of CECP is not very useful / needlessly complicated because it is actually the 'product' of a number of independent or very weakly coupled simpler state machines. There is the ponder state, which is just on or off, affected only by the easy and hard commands...
by hgm
Mon Apr 15, 2019 4:45 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Protocol names of chess variants
Replies: 2
Views: 244

Re: Protocol names of chess variants

I was even thinking to have two comboboxes for this (so that if the engine supports 6 different armies you would not need 36 items in the combobox, but just 2 comboboxes with each 6. Like o normal o FRC o xiangqi o shogi o CwDA: [combo1] - [combo2] FIDE Clobberers Nutters Rookies Dragons Bozos