Alexander Wagner a écrit : > pgeorges schrieb: > > Hi! > > > I uploaded a rc3 of Scid 3.6.19 (with a windows setup). > > Some comments (I used the source, compiling on Debian Etch): > > - Analysis: It would be worth considering to run the > analysis of a game from then end to the start. This allows > the engine to reuse internal hash tables. Stefan told me > that at least within Shredder this improves the analysis > quite a bit as Shredder does not clear the hash tables > without need (aka "not enough RAM"). I guess that other > engines handle this the same way. > Hi, Currently engines don't clear their hash tables when moving forward. But maybe hash data have a higher hit rate when analysing backward. I will check this. > - ./configure: it would be nice to implement --prefix > Any volunteers ? > - Repertoir Window: There exists an old bug when resizing > the window. The empty grey space on the lower end grows > from a certain height onwards. I noticed this first in > 3.6.1 but it might even be older. > > I personaly don't get any "empty grey space" anywhere in the repertoire editor window. Anyway I never use the repertoire feature of Scid : I prefer books and bases, so maybe I missed something here. > > - books' path is now configurable through ./configure and > > Makefile.conf (SHAREDIR) (thanks to Marcin Kasperski) > > - Would be nice to have this settable via the GUI on > userlevel. Consider installing to /usr/local or something > for more than one user. The current setup would require > write permissions to /usr/share/scid/books. > > The same applies for bases subdir. > > Note: ~/.scid/bases or ~/.scid/books might be a better > location but also not be the perfect place so the user > might wish to set these paths manually. > > Ok, that's done now. > > - some code added to support Novag's chess board. This > > code is not finished (and will not be finished until I put > > a hand on a Novag Citrine), so the menu entries are hidden > > to the user until further notice. > > - I'd suggest to implement Novag support similar to dgtdrv > (see http://dgtdrv.sourceforge.net) as then it could be > "sourced out" and you could concentrate on the interface > and limit specific code within scid to a minimum. The > modularisation of dgtdrv is very similar to a chess engine > (that's why I like to call it "input engine"). The same > interface could also be used for DGT and whatever other > board might come without the need to touch scid again. > One could also e.g. implement some TCP/IP or RS232 > connection or whatever by means of this type of protocol. > Also without touching scid. Additionally it keeps the > compile time requirements low and excludes potentially > platform specific code. > > dgtdrv's protocol is very close to uci/xboard so > implementation should be straight forward. I'd guess that > you've already all parts at hand. A proof of concept for > the implementation within xboard by means of a simple > patch is on the page above. I'd volunteer to assist on the > dgt-side, if some extensions to the protocol would be > necessary this would pose no problem. Actually I'm looking > for an implemetation in a real world app to perfectionise > the protocol. > > I prefer to code this in Tcl, not in C or C++. That way compiling is far easier and as there is no performance drawback, coding and debugging is far easier. But I don't have any DGT or Citrine boards, so for now I can't do anything myself. > > - Merida piece sets reworked and added the possibility to > > use image files for white and black squares. See > > wsquare.gif and bsquare.gif in bitmaps directory (Thanks > > to Uwe Klimmek) > > - The entry fields for the bitmaps lack some explaining > text. (Just some "dark squares" and "light squares" in > front of the input line.) > > Done. > - A suggestion would be to use images for the pieces also. > In case I'd have some, IMHO pretty nice, Merida pieces > made from a free TrueType font. You could find them on my > Shredder page > http://theorie.physik.uni-wuerzburg.de/~arwagner/chess/shredder.html > > In case of interest I could also set up a small collection > of bitmaps usable for board designs. > If you have some nice piece sets, you can send them I will incorporate them in the file bitmaps.tcl (I prefer to embed data in tcl, and avoid the use of external files for resources used by Scid).
Pascal ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Scid-users mailing list Scid-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scid-users
