Author: Russell Reagan
Date: 16:10:52 07/31/04
Go up one level in this thread
On July 30, 2004 at 14:42:23, Steven Edwards wrote:
>Also, it uses coordinate move notation instead of SAN,
>and this seems like a ten year step backward.
Do you also use a shotgun to kill flies? I am very surprised to see you saying
this, being the creator of data standards that you are. SAN is major overkill
for use in a communication protocol. Coordinate notation is a much better
choice. It is trivial to support, while SAN is definitely not. Any kid in a
junior high introduction to programming class could support converting
coordinate notation to an internal representation and vice versa. Adding the
same support for SAN will require either an awful piece of error-prone code, or
writing a full blown chess engine. Either way, compared to coordinate notation,
it will be much more error prone and much more time consuming.
Am I missing some redeeming quality that would justify using SAN instead of
coordinate notation for a communication protocol? Please don't say readability,
because that simply isn't an advantage here. If you're reading a chess book, SAN
is more readable. Not so in a communication protocol where A) the moves are
crammed in the middle of all kinds of other unrelated information, and B) the
only time you will be reading the communications is when there is a problem, in
which case you will usually be searching for a known string ("Why isn't the GUI
accepting my engine playing e2e4?") Using SAN doesn't make this any easier: grep
-C 5 'e2e4'
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.