REBEL 16.3

It is also good to study the code of Fruit. Fabian was a true master in the use of assert().

Peter Berger wrote:

On my computer the following position is exceptionally reliable at crashing Rebel 16.3 in infinite analysis after like one hour at most more or less regardless of settings:

[puzzle]
[SetUp  "1"]
[FEN "r2q3r/4bk1p/3p1nbB/1B2p3/pp2P3/6P1/PPPQ4/R1N1K2R w KQ - 0 1"]
[/puzzle]

FEN without graphics: r2q3r/4bk1p/3p1nbB/1B2p3/pp2P3/6P1/PPPQ4/R1N1K2R w KQ - 0 1

If no one can reproduce this, I am out of testing ideas for now.

Kind regards
Peter

Hi Peter,

I tried this position many times with Rebel and Chess System Tal and indeed it stops calculating every time, it may happen 30 minutes or several hours. This is the kind of nightmare bugs. Thanks for reporting.

Admin wrote:

Hi Peter,

I tried this position many times with Rebel and Chess System Tal and indeed it stops calculating every time, it may happen 30 minutes or several hours. This is the kind of nightmare bugs. Thanks for reporting.

Hi Ed,
well, I have to admit to be quite happy, that you get this, too.
What can I add?
1. The position doesn’t crash Rebel 16.2, so there should have been a relevant change (in search?) between this version and Chess System Tal 2.0, that is somehow connected to this bug.
(2. Rebel 16.2 does in fact show kind of similar behaviour at +very+ rare occurences btw, something like once every two weeks or so, but there haven’t been positions that crash it multiple times (just to be complete)).
3. I haven’t seen crashes in endgames or in late middlegame positions yet.
4. This should mean, that in the development of a game there may be a point where Rebel 16.3 starts showing instability and another one where it stops with this unwanted behaviour.
So it may be useful to offer the complete PGN leading to this position:
[Event "bla"]
[Site  "whatever"]
[Date "2024.02.05"]
[Round "-"]
[White "Rebel/CSTal"]
[Black "XXX"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6
6. f3 e5 7. Nb3 Be6 8. Be3 Be7 9. Qd2 Nbd7 10. g4 b5
11. g5 b4 12. Ne2 Nh5 13. Ng3 Nxg3 14. hxg3 a5 15. f4 g6
16. Bd3 a4 17. Nc1 f6 18. f5 Bf7 19. gxf6 Nxf6 20. Bb5+ Kf8
21. fxg6 Bxg6 22. Bh6+ Kf7

I can add, that if we add the obvious 23. Rf1 Bxe4 and let Rebel think about this one, it continues to behave in an unstable way.

Thanks you took the time to have a look.
Peter

Meanwhile I have traced the bastard, this one will not be easy.