Scid 3.6.11 released

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

ikalel

Re: Scid 3.6.11 released

Post by ikalel »

Finally someone took up the job of improving this great piece of software. It was badly needed.

Keep up the good work, man!!
User avatar
Kirk
Posts: 5699
Joined: Sat Mar 11, 2006 3:44 am

Re: Scid 3.6.11 released

Post by Kirk »

Thanks! I appreciate it!
“He knew all the tricks, dramatic irony, metaphor, pathos, puns, parody, litotes and... satire. He was vicious”
User avatar
mariaclara
Posts: 4186
Joined: Wed Mar 08, 2006 9:31 pm
Location: Sulu Sea

Re: Scid 3.6.11 released

Post by mariaclara »

:D :D

hey, am a big fan more than 4yrs now :!: :!: :!:

.....................................Maraming Salamat Po. :!: :!:
(many thanks)

maria

ps

Hi Kirk

:wink: :wink: :wink:
.
.

................. Mu Shin ..........................
pgeorges

Re: Scid 3.6.11 released -> Scid 3.6.12

Post by pgeorges »

I had to make a quick release today :cry: (Scid 3.6.12) because I broke Crafty analysis (which does not affect UCI engines). In the meanwhile I made those changes :

- added new spelling and rating files (thanks to Franz Nagl)
- fixed a bug in opening report that would crash (floating point exception due to a division by 0) if all games in base had a score set to *
- updated swedish translation (thanks to Hans Ericson)
- added a "coach is watching you" in serious game (will detect dubious/weak/bad moves on the fly)
- added a "random level opponent" in tactical games
- fixed a bug in Crafty handling (regression introduced in 3.6.11)
User avatar
fern
Posts: 8755
Joined: Sun Feb 26, 2006 4:07 pm

Re: Scid 3.6.11 released

Post by fern »

Great database. It is full of features. and it works as a charm.
Fdo
User avatar
slobo
Posts: 2331
Joined: Mon Apr 09, 2007 5:36 pm

Re: Scid 3.6.11 released

Post by slobo »

pgeorges wrote:Hi,

I released Scid 3.6.11 at http://prolinux.free.fr/scid/

The main changes are :
- the possibility to play a "serious game" starting from any position and with more time controls (fixed depth, nodes, etc.)
- the management of variations is easier (contextual menu)
- a timer is added to tactical exercises
- added statistics on repertoire knowledge and a report on repertoire knowledge (opening trainer feature) -> I think the opening trainer is useful now ...

Pascal Georges
I admire people like you, Pascal.

Slobo
shiv
Posts: 351
Joined: Sat Apr 01, 2006 2:03 am

Re: Scid 3.6.11 released -> Scid 3.6.12

Post by shiv »

Thanks for the fixes. Quick feedback:
Rather than simply mentioning that the coach thinks that you have blundered, it might be better to have an option (before the game start) which indicates the thresholds to look for. Then depending on the option the user selects, either the wrong move and the right move along with the difference in scores can be annotated on the PGN screen or simply be passed to the user.

Example: The coach thinks you have blundered. Score in position after optimal move is +1.5. Score after your move is -1. Would you like to take your move back?

or the above message is stored in the PGN window as a comment. In essence, the game would be blunderchecked live.
ikalel

Re: Scid 3.6.11 released -> Scid 3.6.12

Post by ikalel »

Example: The coach thinks you have blundered. Score in position after optimal move is +1.5. Score after your move is -1. Would you like to take your move back?

or the above message is stored in the PGN window as a comment. In essence, the game would be blunderchecked live.
I think after you get the coach's warning that you've just played a bad move it would be better to simply append the coach message to the SCID or PGN file with the two numerical evaluations of the position: after the best move and after the move it was actually played.

Revealing the exact numerical evaluation of the position while the game is still in progress is like giving away too much, don't you think? Or, it could be configurable, that is, the user could choose whether he wants to see the numerical evaluation in the popup message or not.
pgeorges

Re: Scid 3.6.11 released -> Scid 3.6.12

Post by pgeorges »

shiv wrote:Thanks for the fixes. Quick feedback:
Rather than simply mentioning that the coach thinks that you have blundered, it might be better to have an option (before the game start) which indicates the thresholds to look for. Then depending on the option the user selects, either the wrong move and the right move along with the difference in scores can be annotated on the PGN screen or simply be passed to the user.

Example: The coach thinks you have blundered. Score in position after optimal move is +1.5. Score after your move is -1. Would you like to take your move back?

or the above message is stored in the PGN window as a comment. In essence, the game would be blunderchecked live.
For now I prefered to make the user interface as simple as possible. Depending on the score drops, the player will get the warnings :
# minimal value for ?!
0.5 -> dubious move
# minimal value for ?
1.5 -> weak move
# minimal value for ??
3.0 -> Bad move

Those values are the same than in "tactical game" and are statically defined for now, but of course the meaning for dubious/weak/bad moves may differ between engines and people. But I prefer to get a message like "you played a dubious move, would you like to take back ?" than a message with numerical values, which I find less human-friendly.
pgeorges

Re: Scid 3.6.11 released -> Scid 3.6.12

Post by pgeorges »

I think after you get the coach's warning that you've just played a bad move it would be better to simply append the coach message to the SCID or PGN file with the two numerical evaluations of the position: after the best move and after the move it was actually played.
As this is supposed to be a "serious game" I prefer not to write down in PGN any kind of analysis as the engine may itself be wrong (usually I give Toga only a few seconds ... and it easily wins). Annotation is a post-mortem process where the engine should analyse at full strength (and not with a limited number of nodes or fixed depth). So during this phase I prefer not to write any analysis, but the user, if taking back a move, will have the opportunity to enter a new variation, and he will see where there were a potential problem.
Revealing the exact numerical evaluation of the position while the game is still in progress is like giving away too much, don't you think? Or, it could be configurable, that is, the user could choose whether he wants to see the numerical evaluation in the popup message or not.
As I already wrote, I prefer human friendly messages instead of numerical scores (which meaning differ a lot between engines : see Rybka and Toga for example).