Critter 1.6a
Moderators: hgm, Rebel, chrisw
-
- Posts: 582
- Joined: Wed May 10, 2006 7:28 pm
- Location: Birmingham, England
Re: Critter 1.6a
Unfortunately this new Critter still crashes on my 64-bit Linux system under Scid when using tablebases. (It's fine otherwise.)
When tablebase usage is set to Everywhere then (with the endgame in sight) it quickly crashes: The analysis engine terminated without exit code:""
When tablebase usage is set to Root then (in an endgame) the analysis window doesn't refresh - so the old analysis remains even when a game is stepped-through and new analysis added. It seems odd in other ways too.
When tablebase usage is set to Everywhere then (with the endgame in sight) it quickly crashes: The analysis engine terminated without exit code:""
When tablebase usage is set to Root then (in an endgame) the analysis window doesn't refresh - so the old analysis remains even when a game is stepped-through and new analysis added. It seems odd in other ways too.
Marek Soszynski
-
- Posts: 1539
- Joined: Thu Mar 09, 2006 2:02 pm
Re: Critter 1.6a
Hello Richard (In case you are reading this)
Do you see any need to retest 1.6a if the 1.6 was already tested? Do you expect any difference in playing strength or are the fixes just "cosmetic"?
How about substituting 1.6 with 1.6a from now on?
Thanks and bye
Ingo
Do you see any need to retest 1.6a if the 1.6 was already tested? Do you expect any difference in playing strength or are the fixes just "cosmetic"?
How about substituting 1.6 with 1.6a from now on?
Thanks and bye
Ingo
-
- Posts: 481
- Joined: Thu Apr 16, 2009 12:00 pm
- Location: Slovakia, EU
Re: Critter 1.6a
This is a bugfix release. The following bugs were fixed:EmilV wrote:http://www.vlasak.biz/critter/
1) engine freezing with dual-socket systems
2) same line displayed 2x in multipv mode
3) occasional crashes with tablebases
4) proper range checking of the depth parameter in "sf store" command
5) linux version now runs on older distributions
as a bonus there is a new UCI option for king safety weight.
Unfortunately I cannot guarantee the ELO performance being same as v1.6. Bug #2 was deep inside the move generator code and caused some moves to be searched twice (not just at the root, but everywhere in the search tree) with an associated performance hit.
-
- Posts: 481
- Joined: Thu Apr 16, 2009 12:00 pm
- Location: Slovakia, EU
Re: Critter 1.6a
Weird. I'm using Scid 4.3 on 64bit Ubuntu 12.04 and was not able to reproduce the crash. Neither the problem with analysis window not refreshing.Marek Soszynski wrote:Unfortunately this new Critter still crashes on my 64-bit Linux system under Scid when using tablebases. (It's fine otherwise.)
When tablebase usage is set to Everywhere then (with the endgame in sight) it quickly crashes: The analysis engine terminated without exit code:""
When tablebase usage is set to Root then (in an endgame) the analysis window doesn't refresh - so the old analysis remains even when a game is stepped-through and new analysis added. It seems odd in other ways too.
-
- Posts: 4567
- Joined: Sun Mar 12, 2006 2:40 am
- Full name:
Re: Critter 1.6a
Also in the Shredder GUI you could see sometimes that: two times the same move in the movelist, as Richard describes it was not in MultiPV mode, I saw it in Analysis mode but the way Richard describes it it probably was in game mode as well . It is not very likely that was optimal for performance, it may not have hurt much but I think that is a bugfix that would require some new testing...IWB wrote:Hello Richard (In case you are reading this)
Do you see any need to retest 1.6a if the 1.6 was already tested? Do you expect any difference in playing strength or are the fixes just "cosmetic"?
How about substituting 1.6 with 1.6a from now on?
Thanks and bye
Ingo
Regards, Eelco
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
-
- Posts: 6442
- Joined: Tue Jan 09, 2007 12:31 am
- Location: PA USA
- Full name: Louis Zulli
Re: Critter 1.6a
Although this will force testers to retest the engine, it's hard to imagine the bug fix resulting in a decrease in ELO.rvida wrote:
Unfortunately I cannot guarantee the ELO performance being same as v1.6. Bug #2 was deep inside the move generator code and caused some moves to be searched twice (not just at the root, but everywhere in the search tree) with an associated performance hit.
Thanks again for your work, Richard.
-
- Posts: 9773
- Joined: Wed Mar 08, 2006 8:44 pm
- Location: Amman,Jordan
Re: Critter 1.6a
Works fine now
What are the estimated values of the king safety option
I mean the range of values
What are the estimated values of the king safety option
I mean the range of values
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
-
- Posts: 1971
- Joined: Wed Jul 13, 2011 9:04 pm
- Location: Madrid, Spain.
Re: Critter 1.6a: range of values of king safety weight.
Hello Dr.:
I do not know if increasing king safety weight makes Critter play more safer/conservative, and decreasing this weight just the opposite... thanks in advance for the possible answers.
Thank you very much for your effort, Richard! Good luck.
Regards from Spain.
Ajedrecista.
I use Arena 2.0.1: in the engine configuration window, the range is from 0 to 100, being 32 as default (the same default value as many weights in Houdini, when HoudiConfig tool is used).Dr.Wael Deeb wrote:Works fine now
What are the estimated values of the king safety option
I mean the range of values
I do not know if increasing king safety weight makes Critter play more safer/conservative, and decreasing this weight just the opposite... thanks in advance for the possible answers.
Thank you very much for your effort, Richard! Good luck.
Regards from Spain.
Ajedrecista.
-
- Posts: 9773
- Joined: Wed Mar 08, 2006 8:44 pm
- Location: Amman,Jordan
Re: Critter 1.6a: range of values of king safety weight.
Thanks for the info....Ajedrecista wrote:Hello Dr.:
I use Arena 2.0.1: in the engine configuration window, the range is from 0 to 100, being 32 as default (the same default value as many weights in Houdini, when HoudiConfig tool is used).Dr.Wael Deeb wrote:Works fine now
What are the estimated values of the king safety option
I mean the range of values
I do not know if increasing king safety weight makes Critter play more safer/conservative, and decreasing this weight just the opposite... thanks in advance for the possible answers.
Thank you very much for your effort, Richard! Good luck.
Regards from Spain.
Ajedrecista.
I think Richard can shed more light related to this issue....
Dr.D
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….