Page 1 of 2

Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 12:46 pm
by Engin
Hello,

why we make differenz between chess960 and standard chess, who as allready include standard chess position #518 ?

we can also chess calling for one as chess960.

and we can do the notation allways with a startposition in XFEN (HAha), and notate allways short castle with O-O and long castle with O-O-O.

I see no problems to change the GUIs only for chess960 with this two rules, one for the startpositions and for the castles.

The books we can also call with the startpositon number (#1 - #960) that include #518 book

why we make castles different with "king takes rook" in chess960 and in stardard chess with e1g1 or e1c1 ?
we are allways know that king is in the mid of the both rooks, and the short castle is for white right side and the long is left side, the rules a same with standard chess, or we can for computers allway say king and rook squares are castles.

The feature chess is for me NOW with Chess960 for all positions !

if we play stardard chess we are playing #518 in chess960, why makes differenz ?


best wishes,
Engin

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 12:59 pm
by hgm
The reason is that you have to remain backward compatible with existing Chess engines that do not understand O-O is casting, but only accept e1g1, or none of those engines would work anymore.

Likewise with FENs: most existing engines that only pay normal Chess and not FRC would not understand HAha castling rights. Even as it is, there are some engines that do not understand HAha casting rights even in FRC... :lol:

In fact KxR is an inferior notation for castling, and it woud be better to use RxK: there might be variants where it is legal to capture your own pieces, and there KxR woud be ambiguous. But even in such variants, it can never be legal to capture your own royal piece.

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 1:08 pm
by Engin
yes, but we can do very easy PGN notations backward compatible with the startposition FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w HAha -" and O-O, O-O-O, but engines are use KxR format for castles

programming chess engines with castle format FEN HAha is i think not a big problem.

And we are not need to switch between Chess960 and Standard Chess, Why?

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 1:26 pm
by Engin
its only need 2 rules make this as standard.

and we can grow our chess horizont.

The normal chess #518 is the most analysed position, but we can starting from now to analyse the other position too. ok they are very much variants and positions, but for computers i think this is not a big problem, to gerate then from PGN different books between #1 - #960 include #518 for normal chess.

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 2:27 pm
by hgm
The problem is that declaring something to be a standard does not make it a standard. What counts is what system is used by that what is actually around. That it can be easily changed is not relevant, as it can be even easier not changed.

Suppose you are a user of a commercial engine that you paid €30 for to get a license to use the executable, and it does not understand O-O as a castling move. Now your favorite GUI decides to declare O-O a standard, and stops supporting e1g1...

How easy exactly is it now for you to change the engine so that it can run under this new GUI?

Btw, there is a FEN standard for Chess960 that is fully backward comptible with FENs for normal Chess (X-FEN). In PGN the standard notation for castling is already O-O in both normal Chess and FRC.

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 6:53 pm
by Engin
even we dont need the FEN castle format HAha, the standard KQkq is enough and in PGN we notate only the startposition, the O-O and O-O-O is same as before, and i found new idea with handle castles moves, that there engines does not play FRC using king to g1/g8 or king to c1/c8 notation, example e1g1/e1c1/e8g8/e8d8 continue, only FRC engines are change a little to use for other FRC positions, example if the king is on d1 use with d1g1/d1c1 or d8g8/d8c8 for castles, its not needed for king takes rook moves. And we dont need to switch between chess960 and standard chess.

GUIs are not much changed and all FRC engines stay backward compatible.

FEN with HAha and king takes rook confuse all things.does not need i think, we can stay old format with only startposition defination.

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 6:56 pm
by MattieShoes
Winboard doesn't support FRC anyway, does it? I was going to write a FR "variant" engine but I was under the impression that it's not supported anyway...

If it does or will, could anybody shed some light on what format it uses? I assume HAha is for castling permissions left/right in the FEN so it's not confused with the side the queen actually started on, but the other part...

(In suicide chess, it's perfectly acceptable to capture a king, or promote a pawn to a king, but you can't capture your own pieces OR castle so...

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 6:59 pm
by Engin
ok this is not new idea, i mean we dont need very much to change between chess960 and standard chess, only to know wich startposition to starting.

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 7:03 pm
by Engin
winboard F does support FRC with option /variant=fisherrandom

but all this confuse variants does not need, we need only the startposition definition.

Re: Chess960 and Standard Chess differenz

Posted: Thu Apr 16, 2009 7:19 pm
by Matthias Gemuh
Engin wrote:even we dont need the FEN castle format HAha, the standard KQkq is enough and in PGN we notate only the startposition, the O-O and O-O-O is same as before, and i found new idea with handle castles moves, that there engines does not play FRC using king to g1/g8 or king to c1/c8 notation, example e1g1/e1c1/e8g8/e8d8 continue, only FRC engines are change a little to use for other FRC positions, example if the king is on d1 use with d1g1/d1c1 or d8g8/d8c8 for castles, its not needed for king takes rook moves. And we dont need to switch between chess960 and standard chess.

GUIs are not much changed and all FRC engines stay backward compatible.

FEN with HAha and king takes rook confuse all things.does not need i think, we can stay old format with only startposition defination.
There are many problems with what you are proposing.
The problems have been discussed here many times.

Code: Select all

example if the king is on d1  use with d1g1
And if the king is on f1, castle with f1g1 ?

Matthias.