When testing a version of Shatranj I plugged it against Ace 0.1 from Adam Adair and noticed again that Ace always moves immediately.
I never had it tested for the RWBC classes, because it appeared too late for them. Yet I had it play in a few rapid tournaments for my old rating list
and I now remembered that symptom(confirmed in other posts here).
It even has an uci option 'Search Depth', but it does nothing here.
(values are impractical anyway: min=0 default=0 max=20)
A check of the still available sources revealed that despite that option
the search depth appears to be fixed at depth 5!
I guess this is the reason for its behaviour.
If someone wants to try to fix it, the sources are here:
http://code.google.com/p/ace-chess/
Ace 0.1 search depth / move times
Moderator: Ras
-
Guenther
- Posts: 4718
- Joined: Wed Oct 01, 2008 6:33 am
- Location: Regensburg, Germany
- Full name: Guenther Simon
-
Guenther
- Posts: 4718
- Joined: Wed Oct 01, 2008 6:33 am
- Location: Regensburg, Germany
- Full name: Guenther Simon
Ace 0.1 new compilation - anyone?
bump!Guenther wrote:When testing a version of Shatranj I plugged it against Ace 0.1 from Adam Adair and noticed again that Ace always moves immediately.
I never had it tested for the RWBC classes, because it appeared too late for them. Yet I had it play in a few rapid tournaments for my old rating list
and I now remembered that symptom(confirmed in other posts here).
It even has an uci option 'Search Depth', but it does nothing here.
(values are impractical anyway: min=0 default=0 max=20)
A check of the still available sources revealed that despite that option
the search depth appears to be fixed at depth 5!
I guess this is the reason for its behaviour.
If someone wants to try to fix it, the sources are here:
http://code.google.com/p/ace-chess/
According to the sources 'SearchDepth' is used instead of the uci option 'Search Depth'. But even then it is fixed to max depth 5 in the available exe. (the uci option probably was planned for further releases)
Of course it must have a few big search bugs too, still it would be nice if someone could compile a version with a higher depth.
sources:
http://code.google.com/p/ace-chess/[/quote]
-
cdani
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Re: Ace 0.1 new compilation - anyone?
[/quote]Guenther wrote:bump!Guenther wrote:When testing a version of Shatranj I plugged it against Ace 0.1 from Adam Adair and noticed again that Ace always moves immediately.
I never had it tested for the RWBC classes, because it appeared too late for them. Yet I had it play in a few rapid tournaments for my old rating list
and I now remembered that symptom(confirmed in other posts here).
It even has an uci option 'Search Depth', but it does nothing here.
(values are impractical anyway: min=0 default=0 max=20)
A check of the still available sources revealed that despite that option
the search depth appears to be fixed at depth 5!
I guess this is the reason for its behaviour.
If someone wants to try to fix it, the sources are here:
http://code.google.com/p/ace-chess/
According to the sources 'SearchDepth' is used instead of the uci option 'Search Depth'. But even then it is fixed to max depth 5 in the available exe. (the uci option probably was planned for further releases)
Of course it must have a few big search bugs too, still it would be nice if someone could compile a version with a higher depth.
sources:
http://code.google.com/p/ace-chess/
Hi. I tried but is not a finished engine. It has basic things to do, and despite increasing the depth to 100 it plays instantly because is not a functional engine.
Daniel José -
http://www.andscacs.com
-
Guenther
- Posts: 4718
- Joined: Wed Oct 01, 2008 6:33 am
- Location: Regensburg, Germany
- Full name: Guenther Simon
Re: Ace 0.1 new compilation - anyone?
Ok, thanks for your attempt Daniel!cdani wrote:Guenther wrote:bump!Guenther wrote:When testing a version of Shatranj I plugged it against Ace 0.1 from Adam Adair and noticed again that Ace always moves immediately.
I never had it tested for the RWBC classes, because it appeared too late for them. Yet I had it play in a few rapid tournaments for my old rating list
and I now remembered that symptom(confirmed in other posts here).
It even has an uci option 'Search Depth', but it does nothing here.
(values are impractical anyway: min=0 default=0 max=20)
A check of the still available sources revealed that despite that option
the search depth appears to be fixed at depth 5!
I guess this is the reason for its behaviour.
If someone wants to try to fix it, the sources are here:
http://code.google.com/p/ace-chess/
According to the sources 'SearchDepth' is used instead of the uci option 'Search Depth'. But even then it is fixed to max depth 5 in the available exe. (the uci option probably was planned for further releases)
Of course it must have a few big search bugs too, still it would be nice if someone could compile a version with a higher depth.
sources:
http://code.google.com/p/ace-chess/
Hi. I tried but is not a finished engine. It has basic things to do, and despite increasing the depth to 100 it plays instantly because is not a functional engine.