Author: Reinhard Scharnagl
Date: 05:24:25 07/23/05
Go up one level in this thread
On July 23, 2005 at 06:31:07, F. Huber wrote: >On July 23, 2005 at 05:20:32, Reinhard Scharnagl wrote: > >Ok, then let´s have a look on SMK´s statements: > >1)"Those extensions have not been to my taste": >That´s of course his _opinion_, which BTW most others (incl. me) are sharing. >However: that´s not a ´presupposition for´ but simply a ´conclusion from´ his >next statements. Such a 'conclusion' has not been founded by arguments. >2)"The extension will cause errors": >Absolutely correct! If any of the existing chess-GUIs (not being prepared for >FRC/Chess960) will receive a "KQkq" castling from a FEN, it will assume kings >on e-file and rooks on a/h-file. But in X-FEN even with this "KQkq" the kings >and rook could be located almost ´anywhere´ - so what do you think the GUI >will do, when getting any castling move (O-O, O-O-O) during the game replay, >without even finding both pieces on their original places?? >In the best case th GUI will simply produce nonsense, in the worst case it may >even crash - both of course are "errors", as mentioned in the above statement! Such errors are not caused by X-FEN. The reason is to try non matching data to a program not written for that. It is also not understanding tic-tac-toe or other game representations. Should those games now been made 'compatible' to Fritz to avoid crashing when trying to load something? The user should know what he is doing. An old version e.g. of MS Word will not understand the latest one. More relevant is to stay compatible to applications already supporting Chess960. But Shredder does not care on existing GUIs and engines, online server and databases already having a lot of CHess960 games stored. >3)"Similiar looking FEN positions will have a different meaning": >Of course this is also only related to castling rights. In X-FEN really _every_ >of the 960 starting positions has the _same_ castling tag "KQkq", although they >have indeed a _completely_different_ meaning (i.e. other positions of king and >rooks, and therefore also an other move sequence). The meaning is clear and identic, because all castling place the involved K+R on to the same positions as ever. "QKqk" are related to just that. Thus it is more natural to use the traditional notation for castling rights. It is not the task of the castling tags to describe pieces' positions, which is already done by beginning of FEN or X-FEN. >4)"Chess960 finally is not compatible to traditional Chess": >Again absolutely correct! Chess960 is a superset of classical chess, and the >big majority (942/960) of the starting positions are definitely _contradicting_ >the classical chess rules (with regard to castling)! >Would you call this ´compatible´?? (If ´yes´, you´ll certainly be the only one) Because here is spoken from a set and its superset, it is obviously that there could be only a downwards compatibility. No serious person claims for a FEN extension, which by itself would make old chess programs suddenly play Chess960. But Chess960 is a superset to Chess and X-FEN full compatible to FEN. >Well, that´s now really my last statement about this annoying X-FEN disussion, >but of course feel free to continue your boring campaign to persuade (or even >try to force) people to use this crazy X-FEN format by spamming almost every >chess forum in this wide world with your X-FEN ´informations´(?). >But wait another half year - then you´ll get answers like this: >"What the hell is X-FEN?" ;-) That might be true, but it is no argument at all concerning compatibility. >(oder auf gut Deutsch: "Welche Hahn wird dann noch nach X-FEN krähen?") So the truth is finally written as: mighty Shredder does what it wants to do, who ever bothers on compatibility and the work which others already have invested. And you are wondering yourself why I am upset? Reinhard.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.