Chess960 and Standard Chess differenz

[hgm]

I agree. Chess960 is not really a different variant from normal Chess. Nevertheless, it is useful for a GUI to know if you want to play FRC or normal, because in FRC you would generally want it to shuffle the postion, why with 'normal' you want a specific initial position, without having to feed it to the GUI as a FEN, or even as a number (518) for every game you start. (This is an even more trick point in 10x8 Chess, where there is not one, ut several standard starting positions, depending on if you play Capablanca, Gothic, Bird, Embassy.)

And even if an FRC-capable engine would not have to know if it is playing FRC or normal, and just await the FEN it gets, it would be useful for the GUI to know if a given engine actually implements the FRC castling rules. Then it can warn the user in advance that the engine does not play FRC, rather than have him discover it when he tries to castle. (Of course this would not be a problem for setups like RBQNKNBR.)

In WinBoard 4.3 I included the option to play any variant as a shuffle game, having the GUI randomly pick an opening setup, or having the user fix an arbitrary setup by giving a number, or asking for a one-time random number hich will be used over and over again. But if you do this for a game that has ordinary castling, the Rooks and King are exempt from the shuffling. So to the GUI this does make variant normal in shuffle mode different from FRC or variant nocastling (the latter two in shuffle mode by default): it randomly picks opening setups in all cses, but from a different set of positions. Note that nocastling can alwayss be played as normal chess. But the GUI still has to know how it should shuffle. This is another reason the distinction is useful.

[Engin]

yes, very easy, even the king is allready on g1, the move is g1g1 !

[hgm]

If the King is on f1, and the Rook on h1, and g1 is empty... How would you know if f1g1 is O-O castling, or a simple King move Kg1?

[Spock]

The problem is that declaring something to be a standard does not make it a standard.

Microsoft would disagree, they do it all the time

[hgm]

And if their market share is 80%, I guess they have a point... But none of us has more than a hand full of engines amongst hundreds, plus a single GUI out of a dozen...

[Matthias Gemuh]

yes, very easy, even the king is allready on g1, the move is g1g1 !

Do you now see that you don't think about the details ?
f1g1 would be either a castle or a simple king move.

Even the KQkq you propose for FRC castle rights will fail in certain situations.


Matthias.

[Engin]

no you are not right, the starting position say where the king is and the rooks. if the king move to f1 and then will castle from f1g1, then this is not a castle move, this is a normal king move.

[Engin]

by the starting position FEN where the king and the rooks are is!

[Engin]

rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq

say where king is and the rooks + if this can castle KQkq
king squares on e1/e8 and rooks squares on a1/a8 and h1/h8 !

if now king is on f1, and then will move to g1, then this cant be castle move.

[Matthias Gemuh]

rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq

say where king is and the rooks + if this can castle KQkq
king squares on e1/e8 and rooks squares on a1/a8 and h1/h8 !

if now king is on f1, and then will move to g1, then this cant be castle move.

Try your logic with this FRC position
[D]rnbqbk1r/pppppppp/8/8/8/8/PPPPPPPP/RNBQBK1R w KQkq -

Is f1g1 a castle or not ?