Need help in Chess 960 support!

[hgm]

OK, as result of the testing: I have added code to make sure that WinBoard's notion of the variant is always reset immediately after a game. The problem seemed to be that in zippy mode, WB initializes the engine (+ board display) early, rather than waiting for a new challenge to come in. But at that time the old variant was still in force, as the new variant was still unkown. I reset the variant to 'normal' now at that point.

How your WinBoard could have been thinking that it was playing 'wild4',while all others were actually playing w22 is still a complete mystery. I added some debug print statements from which we might learn more about it.

Version 4.3.13.h can now be downloaded from the same place as above. I will keep DanaSah frc on line as JokerX for the rest of the evening.

[Roman Hartmann]

Ok, I put Roce online again and will match JokerX as soon as it will show up. Maybe the wild4 is due the polyglot+winboard-combination. No idea ...

Roman

[hgm]

The current alpha (4.3.14i) seems to have solved all ICS problems that occurred when switching between variants. In addition it fixes the bug in the reading of PGNs of FRC games mentioned in the other thread by George.

It reverts to variant normal after every game, as it should, because in zippy mode, the engine is initialized in advance, before it is known what game will come next. And if the variant is not normal then, it would be explicitly set during engine initialization, with a variant command, which later might turn out to be the wrong one, so that another variant command would ne sent. Not every engine might like that.

Btw, version 4.3.14f and later also refrain from sending moves to the engine in zippy mode if the engine does not know the variant of the ICS game we are observing, so you can now observe any wild game without the engine choking on it.

[Edsel Apostol]

I'm currently studying this:
http://chesstigers.de/alte_daten/chesst ... /rules.htm

I noticed that no matter where the rook is as long as it is in the right side of the king and the king is not in the a file, they all end up in g1, g8 for the king and f1, f8 for the rook. Its the same with the left side.

I think it would be easy to implement it now as I'm starting to understand it. Good luck to Joker80 and thanks for the feedback!