2008/9/15 Alexander Wagner <[EMAIL PROTECTED]>
> Hi!
>
> I'm currently playing arround with the annotation feature of
> Scid. Some comments here:
>
> - I'd find it usefull if I could just add the engines
> evaluation values (without the actual lines). This just
> results in a "generate score graph" function.
>
> E.g. in games where I added my own variations already, I'd
> prefer a more manual addition of lines, but the score
> graph is generally a useful thing for blunder checking.
This means another option in an overcrowded dialog box (that even does not
fit in low resolution screens).
Is it a really useful thing ? But feel free to add it if you want (maybe
using notebooks widget in Tk 8.5).
>
>
> - To have a "full" score graph, adding the engines evaluation
> even if the best line is played (by the engines thinking)
> would be helpful. One probably might want to add
> engine-lines only in case of "not best line played".
I thought those options already exist !?
>
>
> - In automatic analysis, I think, the engine should stop
> after the game(s) is/are finished. This would allow my
> notebook to go to sleep after it has finished instead of
> running at 100% CPU usage with the fan at the max in the
> "analyse over night" use. ;)
That's right, but in some cases the user may want to annotate the game up to
the last move, and let it analyze the last position. So this would mean yet
another option ...
>
>
> - Annotate several games: besides specifying a number, it
> might be a very good thing to annotate the games in the
> filter. (E.g. if you have ongoing games in a base and you
> want to auto annotate those finished only.)
Another option ...
>
>
> - NAG values are currently added on the basis of the current
> positions evaluation and not according to the difference
> to the position before the current move. This results in
> a sequence of moves all getting e.g. "=" or "+/-" or
> whatever. I feel, only the move that led to equality or
> superiority should get the NAG value attached to it. (Is
> the current behaviour actually a bug or a feature?)
A feature.
>
>
> - Considering exporting the games to some paper version it
> might be a good idea to add a D NAG if a larger change in
> the evaluation is found. Most likely a critical position
> might have been solved here. Probably this correlates
> with the addition of difference based NAGs?
I never use export functions for reviewing games, given Scid is far better
for that. But feel free to add this.
>
>
> - As only best lines are added in auto-annotation mode
> (which IMHO makes perfect sense) the engine should
> probably be set to single-PV mode automatically to get the
> best performance. On the other hand, if it really
> calculates at 5 PV one could also consider adding 5 PV.
> (But I think using this everywhere might clutter up the
> pgn window quite a bit.)
>
> At least if I manually reset multi-PV to 1 and I start
> auto annotation it should stay there.
Except on Pocket PC, Multi PV mode has little impact on performance (after
the first couple of seconds). But if you have any figures...
>
>
> I feel there's a slight "bug" in UCI init code, ie.
> multi-PV is always reset to the value in engine config and
> not the value currently set by the user in the engine
> analysis window. Ie. if I set Shredder to use 3 lines in
> the engine config, fire up the engine, reduce this to 1
> stop and restart it it is reset to 3 lines in auto
> analysis.
Of course, this is the correct behavior. The engine setup is there to be
used. If you want a particular value for MultiPV, engine config is the place
that has the highest priority.
Pascal
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Scid-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scid-users
