Author: Dan Honeycutt
Date: 17:29:00 06/09/04
Go up one level in this thread
On June 09, 2004 at 17:02:30, Russell Reagan wrote: >On June 09, 2004 at 16:30:20, Anthony Cozzie wrote: > >>The e2e4 format is terrible because it is difficult to read for humans (uh, what >>piece did he move?). > >This is not an issue. No one is going to be reading XML files. Or at least, that >isn't the intent. The intent should be that it is easy for everyone to support a >standard that meets our needs, including the shortcomings of existing standards. > >Even if people did read chess XML files, they would have to be a hell of a >blindfold player to know the position after 50 moves, regardless of whether they >know what piece was moved or not. So what is the point in adding complexity to >the standard? > >>Admittedly parsing SAN is not hard (Zappa has a SAN parser) >>but Long algebraic (Nf3xe5) is easier for humans and computers both. > >Nf3xe5... Uh, what piece was captured? ;-) > >I like long algebraic also. When we talk about the difficulty of supporting SAN, >we have to consider relative difficulty. No, it is not difficult to support SAN, >but compared to coordinate notation, it _is_ difficult. Any kid a week into his >first high school programming class could support coordinate notation. Not a >chance for supporting SAN. That would be a semester long (or year long?) project >:-) > >By the way, did you write your own SAN parser, or did you borrow code? If you >wrote your own, does it _really_ support SAN? i.e., would your program read >moves like Lxf7? I don't know if that is really in the SAN standard or not, but >I've seen people from other countries post PGN containing moves like that. With >coordinate notation, there are no such problems with internationalization. SAN not quite - you still have the issue of promotions. But I agree with what you said elsewhere in this thread that for an internal representation readability should not be a primary concern. the only thing I find comfortable to read is pure SAN using PNBRQK - the hardest to parse and language dependant. for an internal representation I like e2e4 or e7e8=4. Dan H. >and long algebraic both have that issue to deal with. To me, that seems like >something the GUI or conversion tool should deal with, not something an engine >programmer should worry about.
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.