SCID5

Discussion of chess software programming and technical issues.

Moderator: Ras

Fulvio
Posts: 396
Joined: Fri Aug 12, 2016 8:43 pm

Re: SCID5

Post by Fulvio »

dangi12012 wrote: Mon Jul 18, 2022 10:09 am Lichess also adds a %eval comment. This can be used to get thousands of cpu hours of SF 14/15 eval of millions of games.
Yes, it is great.
In SCID the evaluation bar use that comments (if there isn't a local engine running).
It also allows to search (Search->Header->PGN contains text)

Code: Select all

eval #
and find the games where a mate was possible
User avatar
WinPooh
Posts: 276
Joined: Fri Mar 17, 2006 8:01 am
Location: Russia
Full name: Vladimir Medvedev

Re: SCID5

Post by WinPooh »

What about having some specific features for computer players? Not all of them have photo and date of birth but they can have "number of cores" or "network version" fields. Is your database suitable for computer chess?
Fulvio
Posts: 396
Joined: Fri Aug 12, 2016 8:43 pm

Re: SCID5

Post by Fulvio »

WinPooh wrote: Thu Jul 21, 2022 12:50 pm What about having some specific features for computer players?
Player informations are stored separately and independently.
A SCID5 database can be think of as a binary PGN: contains all the same games information, but it is a lot faster and also have a smaller size.
wickedpotus
Posts: 153
Joined: Sun May 16, 2021 5:33 pm
Full name: Aron Rodgriges

Re: SCID5

Post by wickedpotus »

Fulvio wrote: Fri Jul 22, 2022 9:31 pm
WinPooh wrote: Thu Jul 21, 2022 12:50 pm What about having some specific features for computer players?
Player informations are stored separately and independently.
A SCID5 database can be think of as a binary PGN: contains all the same games information, but it is a lot faster and also have a smaller size.
Is current scid versions using the SCID5 format or is the intent to implement it in scid versions later?
sam.shapiro
Posts: 1
Joined: Tue May 26, 2020 4:39 am
Full name: Sam Shapiro

Re: SCID5

Post by sam.shapiro »

Hi Fulvio,
thank you very much for your excellent work on scid
Could you please add the option to remove the move from the polyglot book in Book tuning
Fulvio
Posts: 396
Joined: Fri Aug 12, 2016 8:43 pm

Re: SCID5

Post by Fulvio »

wickedpotus wrote: Sat Aug 13, 2022 8:51 pm Is current scid versions using the SCID5 format or is the intent to implement it in scid versions later?
It is already implemented in the current version:
https://sourceforge.net/p/scid/code/ci/ ... ec_scid5.h
but I have not enabled it in the GUI (I want to finish some things in the new engine's analysis window before moving the version to 5.0.0).
User avatar
phhnguyen
Posts: 1524
Joined: Wed Apr 21, 2010 4:58 am
Location: Australia
Full name: Nguyen Hong Pham

Re: SCID5

Post by phhnguyen »

Fulvio wrote: Tue Aug 16, 2022 6:51 am
wickedpotus wrote: Sat Aug 13, 2022 8:51 pm Is current scid versions using the SCID5 format or is the intent to implement it in scid versions later?
It is already implemented in the current version:
https://sourceforge.net/p/scid/code/ci/ ... ec_scid5.h
but I have not enabled it in the GUI (I want to finish some things in the new engine's analysis window before moving the version to 5.0.0).
Any chance for the new SCID5 to support importing from/exporting to OCGDB? The first and the main purpose of OCGDB is for exchanging databases thus I wish SCID5 could support.
https://banksiagui.com
The most features chess GUI, based on opensource Banksia - the chess tournament manager
Fulvio
Posts: 396
Joined: Fri Aug 12, 2016 8:43 pm

Re: SCID5

Post by Fulvio »

phhnguyen wrote: Fri Aug 19, 2022 1:35 pm Any chance for the new SCID5 to support importing from/exporting to OCGDB? The first and the main purpose of OCGDB is for exchanging databases thus I wish SCID5 could support.
Not in the next version for sure (I have already so many things to finish!).
But adding a new codec is pretty easy, so in the future we will see.

However, I have also fixed the last thing that annoyed me with PGNs: extra newline chars inserted in comments!
I thing that every program, including older SCID versions, annoyingly split lines blindly resulting in things like:

Code: Select all

1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. O-O Nxe4 { I just wrote a simple, not even so
long, comment }
This was super annoying because converting a SCID database to a PGN, and then reconverting it back to a SCID database, would produce different databases.

But now that problem is solved, and I have to ask: why someone would exchange databases using OCGDB instead of PGN? PGN offers incredible stability (there are no problems in opening PGNs created 20 years ago) and portability (it is supported by every chess software and it is even possible to open it with just a text editor).
It can only be beaten in compressed file size. Thats why there are sites where you can find other format like Chessbase's cbv ( https://theweekinchess.com/twic ) or SCID ( http://caissabase.co.uk/ ).
Fulvio
Posts: 396
Joined: Fri Aug 12, 2016 8:43 pm

Re: SCID5

Post by Fulvio »

sam.shapiro wrote: Mon Aug 15, 2022 8:30 am Could you please add the option to remove the move from the polyglot book in Book tuning
I honestly do not know much about polyglot books.
It would pretty easy to change the code so that moves that have weight zero are removed.
So I have to ask to the experts: it would be ok to remove that moves? What's the difference between a move that do not exists and a move with weight zero?
Joost Buijs
Posts: 1632
Joined: Thu Jul 16, 2009 10:47 am
Location: Almere, The Netherlands

Re: SCID5

Post by Joost Buijs »

The difference is that a move with weight zero will still be played if there are no other (better) moves available. A move that doesn't exist can never be played. Of course it all depends upon the criteria you set in the program that uses the database.