Progress on Blunder

[algerbrex]

It says, Blunder engine terminated and then the whole of other running games are terminated by cute chess, then cute chess itself terminates.

Hmmm, thanks. I'll have to do some investigating. Perhaps I'll have to release a hotfix, since no point in testing a buggy version that crashes. Time control was 40 moves/2 mins right?

Apologies for not testing more thoroughly.

Yes, 40/120, 40m/2m. No problem, pretty common for brand new engines, I have restarted at 40/10, perhaps the PGN gives a hint.

Thanks. Hopefully so. I've been meaning to add a logging option for Blunder for a while. Now seems like a good time to do so. I've been running a couple of longer time control games myself to see if I can spot any obvious issues, and nothing really has come up, except that every now and then it looks like the the NPS Blunder's reporting is overflowing...which is interesting, to say the least. I'll look into that more.

[Rebel]

It says, Blunder engine terminated and then the whole of other running games are terminated by cute chess, then cute chess itself terminates.

Hmmm, thanks. I'll have to do some investigating. Perhaps I'll have to release a hotfix, since no point in testing a buggy version that crashes. Time control was 40 moves/2 mins right?

Apologies for not testing more thoroughly.

Yes, 40/120, 40m/2m. No problem, pretty common for brand new engines, I have restarted at 40/10, perhaps the PGN gives a hint.

Thanks. Hopefully so. I've been meaning to add a logging option for Blunder for a while. Now seems like a good time to do so. I've been running a couple of longer time control games myself to see if I can spot any obvious issues, and nothing really has come up, except that every now and then it looks like the the NPS Blunder's reporting is overflowing...which is interesting, to say the least. I'll look into that more.

1600 games without problems but..... I saw the below 3-4 times :

Code: Select all

Finished game 1129 (Loki_3.5.0 vs Blunder_7.3.0): 0-1 {Black wins by adjudication}
Warning: Illegal move: Ndb4
Started game 1156 of 1600 (Blunder_7.3.0 vs Jumbo_0.5.3)

Every time Ndb4

However no mention in the PGN as far as I can see.

[algerbrex]

It says, Blunder engine terminated and then the whole of other running games are terminated by cute chess, then cute chess itself terminates.

Hmmm, thanks. I'll have to do some investigating. Perhaps I'll have to release a hotfix, since no point in testing a buggy version that crashes. Time control was 40 moves/2 mins right?

Apologies for not testing more thoroughly.

Yes, 40/120, 40m/2m. No problem, pretty common for brand new engines, I have restarted at 40/10, perhaps the PGN gives a hint.

Thanks. Hopefully so. I've been meaning to add a logging option for Blunder for a while. Now seems like a good time to do so. I've been running a couple of longer time control games myself to see if I can spot any obvious issues, and nothing really has come up, except that every now and then it looks like the the NPS Blunder's reporting is overflowing...which is interesting, to say the least. I'll look into that more.

1600 games without problems but..... I saw the below 3-4 times :

Code: Select all

Finished game 1129 (Loki_3.5.0 vs Blunder_7.3.0): 0-1 {Black wins by adjudication}
Warning: Illegal move: Ndb4
Started game 1156 of 1600 (Blunder_7.3.0 vs Jumbo_0.5.3)

Every time Ndb4

However no mention in the PGN as far as I can see.

Hmm, very interesting, thanks! Good to see it's at least not crashing all the time. I'll definitely be looking more into this. This should give me some ideas of what the issue might be...

At first glance, I thought cutechess was reporting an illegal PV, but I remembered it'll usually state it's a PV move, so Blunder is probably just sending in an illegal move. Which is pretty odd, considering I haven't touched the move generator in months. Oh well, it's a place to start!

[mvanthoor]

1600 games without problems but..... I saw the below 3-4 times :

Code: Select all

Finished game 1129 (Loki_3.5.0 vs Blunder_7.3.0): 0-1 {Black wins by adjudication}
Warning: Illegal move: Ndb4
Started game 1156 of 1600 (Blunder_7.3.0 vs Jumbo_0.5.3)

Every time Ndb4

Hmm, very interesting, thanks! Good to see it's at least not crashing all the time. I'll definitely be looking more into this. This should give me some ideas of what the issue might be...

At first glance, I thought cutechess was reporting an illegal PV, but I remembered it'll usually state it's a PV move, so Blunder is probably just sending in an illegal move. Which is pretty odd, considering I haven't touched the move generator in months. Oh well, it's a place to start!

Maybe I'm reading this wrong, but it says:

Loki vs. Blunder
Black wins by adjudication
Illegal move (which would be the reason for adjudication)

Is Blunder black in this game? If so, it wins and it's Loki making the illegal moves.

[algerbrex]

1600 games without problems but..... I saw the below 3-4 times :

Code: Select all

Finished game 1129 (Loki_3.5.0 vs Blunder_7.3.0): 0-1 {Black wins by adjudication}
Warning: Illegal move: Ndb4
Started game 1156 of 1600 (Blunder_7.3.0 vs Jumbo_0.5.3)

Every time Ndb4

Hmm, very interesting, thanks! Good to see it's at least not crashing all the time. I'll definitely be looking more into this. This should give me some ideas of what the issue might be...

At first glance, I thought cutechess was reporting an illegal PV, but I remembered it'll usually state it's a PV move, so Blunder is probably just sending in an illegal move. Which is pretty odd, considering I haven't touched the move generator in months. Oh well, it's a place to start!

Maybe I'm reading this wrong, but it says:

Loki vs. Blunder
Black wins by adjudication
Illegal move (which would be the reason for adjudication)

Is Blunder black in this game? If so, it wins and it's Loki making the illegal moves.

Hmm, now that you mention it, Marcel, that does seem to be the case. If cutechess is adjudicating in favor of Black, that should mean white loss, and white made the illegal move.

[Guenther]

1600 games without problems :D but..... I saw the below 3-4 times :

Code: Select all

Finished game 1129 (Loki_3.5.0 vs Blunder_7.3.0): 0-1 {Black wins by adjudication}
Warning: Illegal move: Ndb4
Started game 1156 of 1600 (Blunder_7.3.0 vs Jumbo_0.5.3)

Every time Ndb4 :!:

Hmm, very interesting, thanks! Good to see it's at least not crashing all the time. I'll definitely be looking more into this. This should give me some ideas of what the issue might be...

At first glance, I thought cutechess was reporting an illegal PV, but I remembered it'll usually state it's a PV move, so Blunder is probably just sending in an illegal move. Which is pretty odd, considering I haven't touched the move generator in months. Oh well, it's a place to start!

Maybe I'm reading this wrong, but it says:

Loki vs. Blunder
Black wins by adjudication
Illegal move (which would be the reason for adjudication)

Is Blunder black in this game? If so, it wins and it's Loki making the illegal moves.

Hmm, now that you mention it, Marcel, that does seem to be the case. If cutechess is adjudicating in favor of Black, that should mean white loss, and white made the illegal move.

Every time Ndb4 :!:

However no mention in the PGN as far as I can see.

Well look again at the quote snippet from Ed above. The illegal moves actually were not made on the board, otherwise
the games would have been forfeited. (I checked the pgn too and there is nothing like this for the mentioned 1600 games)

I also remember something like illegal pv move messages for cute, but those were mostly along the whole pv, may be cute prints
another error message, if the first move of a pv is already illegal, or the pv ist just one move long and illegal.
A debug file would help more here.

Edit:
Oh looking at the message again, it could also be something very else!
What if Eds start pgn just has a typo!? Then there will be probably such an error message too and the game(s) with taht start position is/are discarded.

[Guenther]

Well look again at the quote snippet from Ed above. The illegal moves actually were not made on the board, otherwise
the games would have been forfeited. (I checked the pgn too and there is nothing like this for the mentioned 1600 games)

I also remember something like illegal pv move messages for cute, but those were mostly along the whole pv, may be cute prints
another error message, if the first move of a pv is already illegal, or the pv ist just one move long and illegal.
A debug file would help more here.

Edit:
Oh looking at the message again, it could also be something very else!
What if Eds start pgn just has a typo!? Then there will be probably such an error message too and the game(s) with taht start position is/are discarded.

pgn parsing illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L200
engine pv illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L833

:)

[algerbrex]

Well look again at the quote snippet from Ed above. The illegal moves actually were not made on the board, otherwise
the games would have been forfeited. (I checked the pgn too and there is nothing like this for the mentioned 1600 games)

I also remember something like illegal pv move messages for cute, but those were mostly along the whole pv, may be cute prints
another error message, if the first move of a pv is already illegal, or the pv ist just one move long and illegal.
A debug file would help more here.

Edit:
Oh looking at the message again, it could also be something very else!
What if Eds start pgn just has a typo!? Then there will be probably such an error message too and the game(s) with taht start position is/are discarded.

pgn parsing illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L200
engine pv illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L833

Ah, thanks Guenther. That looks like that's the issue! I think I'll try to reproduce this myself and see what happens.

If that is the issue then that'll make me feel quite a bit better, since I wasn't looking forward to digging through Blunder's move generation code, which I haven't touched in months

[Guenther]

Well look again at the quote snippet from Ed above. The illegal moves actually were not made on the board, otherwise
the games would have been forfeited. (I checked the pgn too and there is nothing like this for the mentioned 1600 games)

I also remember something like illegal pv move messages for cute, but those were mostly along the whole pv, may be cute prints
another error message, if the first move of a pv is already illegal, or the pv ist just one move long and illegal.
A debug file would help more here.

Edit:
Oh looking at the message again, it could also be something very else!
What if Eds start pgn just has a typo!? Then there will be probably such an error message too and the game(s) with taht start position is/are discarded.

pgn parsing illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L200
engine pv illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L833

:)

Ah, thanks Guenther. That looks like that's the issue! I think I'll try to reproduce this myself and see what happens.

If that is the issue then that'll make me feel quite a bit better, since I wasn't looking forward to digging through Blunder's move generation code, which I haven't touched in months :lol:

No news from Ed?
Easiest way would have been to check his gambit lines, but they are still secret it seems...

[algerbrex]

Well look again at the quote snippet from Ed above. The illegal moves actually were not made on the board, otherwise
the games would have been forfeited. (I checked the pgn too and there is nothing like this for the mentioned 1600 games)

I also remember something like illegal pv move messages for cute, but those were mostly along the whole pv, may be cute prints
another error message, if the first move of a pv is already illegal, or the pv ist just one move long and illegal.
A debug file would help more here.

Edit:
Oh looking at the message again, it could also be something very else!
What if Eds start pgn just has a typo!? Then there will be probably such an error message too and the game(s) with taht start position is/are discarded.

pgn parsing illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L200
engine pv illegal move warning
https://github.com/cutechess/cutechess/ ... e.cpp#L833

Ah, thanks Guenther. That looks like that's the issue! I think I'll try to reproduce this myself and see what happens.

If that is the issue then that'll make me feel quite a bit better, since I wasn't looking forward to digging through Blunder's move generation code, which I haven't touched in months

No news from Ed?
Easiest way would have been to check his gambit lines, but they are still secret it seems...

I suppose so. Or maybe he's just away from the forum for a while. It would make sense with the holidays coming up.