Laser 1.3 beta by Jeffrey An and Michael An.
Roughly 2850 ELO (CCRL)
Regressions vs. 1.1:
+114 in 8000 games at 3+0.02
+109 in 4000 games at 15+0.05
Regressions vs. 1.1:
+114 in 8000 games at 3+0.02
+109 in 4000 games at 15+0.05
Major changes from 1.1:
Improved king safety
Improved passed pawn evaluation
Slight changes in pruning and reduction
Faster compile
Bugfixes
Improved king safety
Improved passed pawn evaluation
Slight changes in pruning and reduction
Faster compile
Bugfixes
Hi Aprijal, how about Gull 3 64-bit,arm 7,arm8x64? IvanHoe 9.46h same builds,Protector 1.9.0 same build's,Gull R375 64 same thing, thank you!
ReplyDeleteAny ,arm7 pgo, arm8 x64? thank you!
ReplyDeleteThat's arm7 pgo, but no difference at nps.
ReplyDeleteAnother issue, from 6 games played against Laser 1.3,in 4 games engine resigns after 1.e4 e5 2.h3 end..
DeletePlus, engine doesn't show any info, kns,e.t.c.
Thats error happen in CfA gui, but it work in DroidFish, can you verify thats?
DeleteYes, working fine in Droidfish, issue in Cfa.
DeleteLaser engine, can actually play in C4a, it need more time to think, for example, 5 sec per move, but still it doesn't show any info-depth, kns, e.t.c.
DeleteV0.2 needed polyglot according to Jim Ablett. The same may be repeating. If so, a specific build of polyglot must be compiled with reference to the config file. That means like for example Laser.ini inside Polyglot binary. So, polyglot looks for ini file and the ini file looks for the engine. CfA uses the polyglot binary as an xboard engine.
ReplyDeleteI've tested the polyglot trick by replacing just the engine file in my current Laser 0.2.JA config. It works with CfA. The theory is valid. I will upgrade Laser in Rapidroid using this workaround.
ReplyDelete