Author: Steven J. Edwards
Date: 23:06:48 05/30/98
Go up one level in this thread
On May 31, 1998 at 01:18:04, Peter Klausler wrote:
>A comment on, well, commentary in PGN games. A deficiency
>with the present Standard is that commentary in movetext
>applies to both the positions reached as well as to the
>context of the game. It's not presently possible to
>distinguish between comments that apply to positions
>independent of whatever game in which they arise
>(such as evaluative scores, principal variations,
>etc.) and those comments that belong only with that
>one game (time trouble, previous moves by the player
>in that position, tournament hall catching fire, etc.)
This is an interesting point. For common game-level items (vs.
move-level or position-level items), the preferred way is to use tag
pairs. There are a number of good reasons for this, two of which are
ease of standardization (documentation, also) and tag value
searchability which is already implemented in (some) PGN DB
applications.
>In CDB, in order to not lose information, I have assumed
>that all comments pertain to positions. This has the
>unfortunate effect of replicating game commentary
>inappropriately when exporting other games containing
>the same positions.
Perhaps user-set options could help with this. Maybe sometimes a user
might want to export the additional information.
The current PGN standard says that comments can be about anything. A
NAG should always refer to the previous move (or previous position).
Same thing with a Broket Form.
>I very much like the idea of permitting EPD descriptive
>annotations in a standardized form to appear in the movetext.
>I would like to suggest that the new <> notation be
>intended to apply to positions, while the old {} notation
>be used for comments that are contextual with the one
>game, including comments on the moves.
>
>Unfortunately, all comments today are in the {} notation,
>and it seems to me that they usually apply to positions
>instead of to games. So I should modify my suggestion
>to be this:
>
> < standardized positional annotations >
> << game-specific commentary, probably not standardizable >>
> { arbitrary comments }
>
>The NAGs presently use a single syntax to convey both
>positional and game information, but I don't think that
>there is ambiguity since their meanings are pretty clear.
>
>A related issue that I have had to deal with involves the
>appearance of commentary prior to movetext, in games as well
>as at the heads of variations. I've handled such comments
>in special ways, and perhaps their semantics should be
>clarified.
Perhaps we can have the convention that any commentary appearing prior
to the first move in the movetext is game specific. This seems to
address most of your concerns without modifications to the syntax.
Remember also that text comments can be both {} style and ; style.
-- Steven (sje@mv.mv.com)
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.