Author: Peter Klausler
Date: 01:11:01 05/30/98
Go up one level in this thread
Some areas where PGN can be strengthened as an interchange
mechanism:
1) Specification of an escape convention to allow the '"'
character in quoted strings and the '{' character in
inline commentary
2) A means of specifying "... with the idea of..." move
sequences
3) A canonical mapping between Informator glyphs and NAGs
4) Integer values for tags (like WhiteElo) that have to be
integers, instead of quoted strings
5) Death penalty for programmers who emit zeroes for castling
6) Making the redundant SetUp tag optional when the FEN tag is
also present
Further, the appearance of a revised PGN standard is an opportunity
to enforce some stricter interpretations in importation routines.
PGN '98 files should be distinguished with a distinct file type
or (better) with some kind of internal marker. This would allow
old sloppy PGN files to continue to be imported, while allowing
new files to be held to a higher standard of adherence to the
spec. I.e., the new PGN definition need not necessarily include
the present language as a compatible subset, since it's easy
enough for an extant parsing routine to handle both. Indeed,
the new PGN language might well benefit from being defined as
a restricted subset of the current language PLUS some syntactically
distinguishable new capabilities; in this way, PGN files emitted
under the new definition that don't use the new capabilities would
still be compatible with the old standard.
I am very enthusiastic about improving PGN (as a compiler writer
by trade, I think it's in pretty good shape already) and will
commit to supporting the new definition in CDB as soon as it
is finalized.
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.