Hello,
I've checked in revision #619. The novelty is a player dictionary
with filtering and sorting functionality.
Gregor
I will repeat here some of my replies from Scid's user forum
about the way of Scidb.
--------------------------------------------------------------------
> ...I am not interested in. No alternative for Chessbase, and
> functionality to manage large databases ...
Yes, ChessBase is designed for huge database, and Scid/Scidb is
definitively not designed for huge database. More than 1.5 million
games in a database is not recommended for Scid/Scidb. ChessBase is
ChessBase, and Scidb will not be ChessBase, Scidb will be Scidb (not
a fork of Scid as well!). Scid/Scidb provides some functionality which
ChessBase cannot provide, for example the very fast position search
(ChessBase is working with memory mapping, no chance; furthermore
ChessBase has to decrypt his game data!). The user has to decide what
he wants, and if Gerd likes to work with huge database then ChessBase
is the primary choice. But if the user is only working with
specialized databases (only the opening repertoire he is usually
playing, for example) he may prefer the features of Scid/Scidb.
--------------------------------------------------------------------
> What I saw last from Scidb there was no database maintenance,
> no filtering, no trees. Instead FICS and PDF-Export seem to be
> a goal…
No, that is definitively not the main goal of Scidb. But developing
software is very, very time consuming, and it needs some time to
implement these things.
For example filtering and views: this is a very important task, but
due to my analysis I cannot implement this without having a tiled
window layout (displaying the results in views). And implementing
this tiled window layout is still in progress. I had to start from
the scratch, the Tk library does not offer any support for this.
The PDF export only a secondary task, and the tiled window layout
will come first.
The support of FICS games is a basic functionality, it belongs
to the database format, and I cannot provide a production version
until the database format is stable. I will have a stable format
this year (after about four years analysis and research). As soon
as the database format is stable the maintenance functions will be
implemented, stability of the format is a precondition.
The main features of Scidb's database format (so far):
1. It is supporting different chess variants, and provides much more
information than Scid's. The load of a Scidb database is up to four times
faster than Scid's load of a database (if Scidb is compiled with optimization
on - the current preview version is not optimized and has a lot of debugging
code; furthermore this depends on the database size and disk speed). In some
cases Scid's database on disk is up to three times larger than Scidb's.
2. Scidb is (fully) supporting some chess variants:
- Chess 960 positions
- Shuffle Chess positions
- Some special FICS start positions
(wild/8a, pawns/little-game, and so on)
- Crazyhouse
- Bughouse (in near future)
- Antichess (Suicide, Giveaway)
- Losers Chess (a variant of Antichess)
- Three-Check Chess
3. Scidb is the first chess database which can import any PGN file
(without any error), for example PGN files from FICS (any variant).
In my opinion this is very important because stability and robustness
is one of the most important goals. Furthermore Scidb's PGN parser is
supporting PGN files from Rybka Aquarium.
4. In next version I will support attachment of documents to a database.
--
Gregor Cramer
email: <gcra...@gmx.net>
