TalkChess.com

Ras wrote: ↑Wed Oct 16, 2019 5:35 pm

Dann Corbit wrote: ↑Tue Oct 15, 2019 1:11 am For very high speed games, sending only the move is a lot better than sending the whole game state including the board.

The moves aren't being sent over a 300 baud line, so "a lot better" doesn't hold water.

Most testing done today is at ludicrously high speed (veiled "Spaceballs" reference). For instance, a Stockfish STC run may consume one quarter million games in just a few minutes.

A chess engine cannot just ignore the extra information because there is no way to know if an incoming board is the next move or if we are now going to analyze a chess position from Fischer verses Tal.

For game play, thé Xboard protocol isn't just better, it's a lot better. Especially for testing, but also for learning.

IMO, YMMV

Dann Corbit wrote: ↑Wed Oct 16, 2019 6:10 pmFor game play, thé Xboard protocol isn't just better, it's a lot better.

Then that massive advantage explains why all of the top engines are CECP engines. Oh, wait...

Robert Pope wrote: ↑Mon Oct 14, 2019 8:46 pm Also, I believe that UCI doesn't send a game result/end game flag to the engine, so if you are doing any post-game processing, you have to infer when this occurs, rather than being told explicitly.

Depending on the circumstances you can coax ucinewgame for this, or just detect when the current position isn't a continuation of the previous one.

Ras wrote: ↑Wed Oct 16, 2019 7:23 pm

Dann Corbit wrote: ↑Wed Oct 16, 2019 6:10 pmFor game play, thé Xboard protocol isn't just better, it's a lot better.

Then that massive advantage explains why all of the top engines are CECP engines. Oh, wait...

^infinity

You probably also think that Windows must be hugely better than Linux?

hgm wrote: ↑Wed Oct 23, 2019 9:31 am You probably also think that Windows must be hugely better than Linux?

It was Willy Wonka who said, "Strike that, reverse it."