booot 7.0 final release

[booot]

Stafan, may be you can send me the debug file (with warnings)?

[pohl4711]

Thanks! I really changed much in this version. I test engine only in Arena and did not see bugs. I will use cutechess-cli to catch them. What conditions did you use for testing (time control, number of threads etc.)?

Alex.

Its all on my website...
But here again:
SPCC Top Engines Ratinglist (+ regular testing of Stockfish Dev-versions)
Playing conditions:
Hardware: Since 20/07/21 AMD Ryzen 3900 12-core (24 threads) notebook with 32GB RAM.
Speed: (singlethread, TurboBoost-mode switched off, chess starting position) Stockfish 14.1: 750 kn/s (when 20 games are running simultaneously)
Hash: 256MB per engine
GUI: Cutechess-cli (GUI ends game, when a 5-piece endgame is on the board)
Tablebases: None for engines, 5 Syzygy for cutechess-cli
Openings: HERT_500 testset (by Thomas Zipproth) (download the file at the "Download & Links"-section or here). Mention, the HERT-set is not an Anti-Draw (UHO or something) opening-set, but a classical, balanced opening-set.
Ponder, Large Memory Pages & learning: Off
Thinking time: 180sec+1000ms (= 3min+1sec) per game/engine (average game-duration: around 7.5 minutes). One 7000 games-testrun takes about 2 days.

[Wolfgang]

Thanks! I really changed much in this version. I test engine only in Arena and did not see bugs. I will use cutechess-cli to catch them. What conditions did you use for testing (time control, number of threads etc.)?

Alex.

No problems here, not a single crash, no time forfeits, no cryptic notifications. Running rock stable as the beta did too.
Mainly with BanksiaGUI, 12-14 games played simultaneously on a 16Core machine.
Thanks Alex

[booot]

Thanks! I really changed much in this version. I test engine only in Arena and did not see bugs. I will use cutechess-cli to catch them. What conditions did you use for testing (time control, number of threads etc.)?

Alex.

No problems here, not a single crash, no time forfeits, no cryptic notifications. Running rock stable as the beta did too.
Mainly with BanksiaGUI, 12-14 games played simultaneously on a 16Core machine.
Thanks Alex

Thank you Wolfgang!
But... seems bad news for me : the mysterious bug is much worse the the stable.

[Wolfgang]

Our tester Jörg uses Cutechess GUI , not "cli".
No problems there too

[booot]

Our tester Jörg uses Cutechess GUI , not "cli".
No problems there too

Hm.... So the bug nest is located. I will take a look to "cli' - i have never used it before.

[booot]

How is results, Wolfgang? Booot traditionally has difficulties in CEGT team

[Wolfgang]

https://cegt.forumieren.com/t1802-testing-booot-7-0nn

My games are finished, will post the results later.
Looks like +60 to v6.5

[Wolfgang]

1200 games played (more to follow): +75 so far

[pohl4711]

Our tester Jörg uses Cutechess GUI , not "cli".
No problems there too

Hm.... So the bug nest is located. I will take a look to "cli' - i have never used it before.

The bug nest is the booot7 engine. Cutechess-cli is a commandline-tool. If an engine does not work there, it is buggy. If other GUIs ignore bogus engine outputs, that does not make the engine working correctly. Just hiding its buggy behaviour.
So, Cutechess-cli should always be the first test-tool for any engine. If an engine runs fine on this deep command-line level, without any GUI (doing weird things on their own), the engine is OK.
Strange thing here is: If I try to start a booot7-test gauntlet on a fresh restarted (offline) PC, some of the 20 booot-instances are terminating (as I wrote before). When I stop cutechess-cli and restart the gautlet immediately, then all 20 booot7 instances seem to run. So, perhaps a memory-allocation bug or so in booot7. Mention, that I tested more than 150 engines/versions in the last 2 years on this PC, using cutechess-cli. Never had such a strange behaviour of an engine until now. The illegal PV-move warnings were caused by some (rare) engines/versions before (first Uralochka for example). But they were all fixed by the engine authors.
Cutechess-cli is not the problem. If it is used for testing by engine-developers, it is the solution for getting a proper working engine in all GUIs!