I've been adding a couple of SCID's newer features, and i guess it's time for a release. The main things are the interruptable tree
processing, support for custom flags, annotation and book changes, and a handy Gamelist speedup.
Tree processing should be on parity with SCID.
The Search and Maintenance widgets now support custom flags, and there are new widgets in the Gamelist to manipulate them also.
The annotation feature has been changed - mainly to make it human useable - and to add a "score only" feature.
The Book window now supports book comparison, and tuning has been redone a little.
There was a couple of bottlenecks in the Gamelist widget, but it handles huge DBs quite nicely now.
Hopefully Ubuntu 11 should compile out of the box too.
One note - I'm not entirely happy with the speed of the Tree's Fast Mode with ECO sorted bases. Pascal's doco claims it should be significantly
faster than Slow Mode, but my testing (with ScidvsPC and Scid) shows otherwise, so ~perhaps~ both projects have a bug here as
there's been some work since Pascal left.
Hi Steven, I need to create a tournament, when I change the number of engines to 4 it show this msg:
Index 3 out of range
Index 3 out of range
while executing
"$l.$i current $i"
(procedure "drawCombos" line 21)
invoked from within
"drawCombos"
invoked from within
".comp.engines.top.update invoke"
("uplevel" body line 1)
invoked from within
"uplevel #0 [list $w invoke]"
(procedure "tk::ButtonUp" line 24)
invoked from within
"tk::ButtonUp .comp.engines.top.update"
(command bound to event)
How does SCID handle Swiss tournaments, and do see any merit in the idea proposed in the ChessGUI thread to let GUIs use an external 'pairing engine' for Swiss tourneys? So that in the future the user would be able to choose which particular flavor of Swiss he wanted to use. (The FIDE handbook already describes 4 different Swiss systems, each of those so complicated that most GUIs that support Swiss tourneys implement their own approximation...)
I decided to give the idea a try in WinBoard, meaning there already is a simple demo pairing engine available. You just send it the number of players, and the tourney results so far (as a string of +-=, in the order of the pairing the engine gave earlier), and after that you can query the engine for the pairing of a specific game in any of the rounds already played, or the round to come.
The hope is of course that we soon will have perfect pairing engines.
tttony wrote:Hi Steven, I need to create a tournament, when I change the number of engines to 4 it show this msg:
Index 3 out of range
Index 3 out of range
while executing
"$l.$i current $i"
(procedure "drawCombos" line 21)
invoked from within
"drawCombos"
invoked from within
".comp.engines.top.update invoke"
("uplevel" body line 1)
invoked from within
"uplevel #0 [list $w invoke]"
(procedure "tk::ButtonUp" line 24)
invoked from within
"tk::ButtonUp .comp.engines.top.update"
(command bound to event)
EDIT: I see, how I install engines??
Yes, this is a little bug.
I think you probably only have the three default engines installed. To install more engines , go to Tools->AnalysisEngines and press F1 for help.
BTW, i like this feature and hopefully next release i'll get round to implementing "whole game" time controls as well as the current "seconds per move".
hgm wrote:How does SCID handle Swiss tournaments, and do see any merit in the idea proposed in the ChessGUI thread to let GUIs use an external 'pairing engine' for Swiss tourneys? So that in the future the user would be able to choose which particular flavor of Swiss he wanted to use. (The FIDE handbook already describes 4 different Swiss systems, each of those so complicated that most GUIs that support Swiss tourneys implement their own approximation...)
I decided to give the idea a try in WinBoard, meaning there already is a simple demo pairing engine available. You just send it the number of players, and the tourney results so far (as a string of +-=, in the order of the pairing the engine gave earlier), and after that you can query the engine for the pairing of a specific game in any of the rounds already played, or the round to come.
The hope is of course that we soon will have perfect pairing engines.
Swiss tournaments are supported by Scid's Crosstable.
Scid vs. PC's Tournament feature (SCID doesn't have one) is fairly basic and certainly doesn't support Swiss tourneys.
I'm not sure there's even a reason to implement this format in a Chess GUI.... Swiss tournaments are sociable things for large groups of people. Chess Engines certainly won't mind being classified into tiers, as their relative strength is known, and they don't suffer from off-days like people.
Well, ChessWar employs a Swiss format. I must admit that I don't know how many other people do. The on-line engine tournaments like CCT, WCRCC and my own monthly blitz tournament are coonducted as Swiss as well, using a tournament manager on the ICS. The open-source TM 'mamer' that I use actually does quite poor pairing when you set if for Swiss. (And we have to, as we have time for 9-10 games, while there are usually at least 14 participants.) It can, for instance, give someone three black games in a row. So mamer could definitely benefit from using a better pairing engine.
Anyway, would you ever consider adding Swiss, the idea is that the pairing engine should make it easy to do so. You wouldn't have to bother implementing all those complcated algorithms for determining who plays who.
Hello and thanks for the nice update to Scid. One problem I can't figure out is when I install Deep Junior 12.5 in only works when the hash tables are set to 16 Mb, anything higher and the engine does not start. This is not a problem with all the other top engines I've installed. Any thoughts would be appreciated.
Best wishes,
Gerald
Amstaff wrote:Hello and thanks for the nice update to Scid. One problem I can't figure out is when I install Deep Junior 12.5 in only works when the hash tables are set to 16 Mb, anything higher and the engine does not start. This is not a problem with all the other top engines I've installed. Any thoughts would be appreciated.
Best wishes,
Gerald
Sorry, but not familiar with this unusual issue. There ~are~ sometime issues with engines, and Scid doesnt get tested much on windows or with commercial products. I had a look at Critter-1.0.1 on WinXP and there was a vaguely reproducible problem associated with it using it's own custom book, but i couldn't pin it down and the thing seemed to go away. Have a look at "Engines (Debugging)" in ScidvsPC's Help index for a little info.
