New engine - Andscacs
Moderators: hgm, Rebel, chrisw
-
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Re: New engine - Andscacs.
Thanks for all your support. Will see how strong is in slow games. I have no clue.
Daniel José - http://www.andscacs.com
-
- Posts: 536
- Joined: Thu Mar 09, 2006 3:01 pm
Re: New engine - Andscacs.
Congratulations on creating a working engine!
Just a couple of suggestions which will make it easier to use and test.
Firstly, it does not run for me on Win7 but it does work on Vista.
You could try building it without needing any MSVCP100.dll or other libs.
Next, it does use a lot of "extra" CPU time, I suspect in your input testing. I am only able to run 2 instances (without pondering) and they consume between 70%-90% of my quad's resources.
Finally (for now), you could add a simple resign capability which would shorten games that are almost certainly lost anyway.
Regards,
Brian
(author of Tinker)
PS Thanks for posting your tuning approach.
Just a couple of suggestions which will make it easier to use and test.
Firstly, it does not run for me on Win7 but it does work on Vista.
You could try building it without needing any MSVCP100.dll or other libs.
Next, it does use a lot of "extra" CPU time, I suspect in your input testing. I am only able to run 2 instances (without pondering) and they consume between 70%-90% of my quad's resources.
Finally (for now), you could add a simple resign capability which would shorten games that are almost certainly lost anyway.
Regards,
Brian
(author of Tinker)
PS Thanks for posting your tuning approach.
-
- Posts: 1221
- Joined: Wed Mar 08, 2006 8:28 pm
- Location: Florida, USA
Re: New engine - Andscacs
Hi Danial,
However, I disagree with your comment that it's "trivial" to create a 2200 ELO engine. When I started to write Maverick I also thought it would be trivial to write a strong engine. But having brought Maverick up to about 2300 ELO I would say it isn't "trivial "to write a decent engine. There are many moving parts; all of which need to be implemented. I think a better description is "straightforward". In other words, if you research what is needed to create a strong engine there is clear path. But you still need to put the time in and code (and debug). This isn't trivial - maybe it's just semantics.
Anyway congrats and I look forward to Maverick locking horns with Andscacs in the future!
Steve
First big congrats on the engine. I've run some quick tests and it does indeed seem to be close to 2400 ELO on the CCRL scale.cdani wrote:(...)The easy conclusion of my last 5 months programming this in some spare hours, is that there is so many information available, that today is near trivial doing a 2200+ engine, and with not much work a more stronger one.
However, I disagree with your comment that it's "trivial" to create a 2200 ELO engine. When I started to write Maverick I also thought it would be trivial to write a strong engine. But having brought Maverick up to about 2300 ELO I would say it isn't "trivial "to write a decent engine. There are many moving parts; all of which need to be implemented. I think a better description is "straightforward". In other words, if you research what is needed to create a strong engine there is clear path. But you still need to put the time in and code (and debug). This isn't trivial - maybe it's just semantics.
Anyway congrats and I look forward to Maverick locking horns with Andscacs in the future!
Steve
http://www.chessprogramming.net - Maverick Chess Engine
-
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Re: New engine - Andscacs.
Thanks!brianr wrote:Congratulations on creating a working engine!
Just a couple of suggestions which will make it easier to use and test.
Firstly, it does not run for me on Win7 but it does work on Vista.
You could try building it without needing any MSVCP100.dll or other libs.
Next, it does use a lot of "extra" CPU time, I suspect in your input testing. I am only able to run 2 instances (without pondering) and they consume between 70%-90% of my quad's resources.
Finally (for now), you could add a simple resign capability which would shorten games that are almost certainly lost anyway.
I will work this weekend on fixing all those things. I have no experience on chess engines and I didn't know exactly if some things I have done are the standards, or if I cover all the expected features.
I will try with every computer I can access to expand the possible configurations where it can run.
Well, yes, better I say it's a lot easier than 10 years agoSteve wrote:However, I disagree with your comment that it's "trivial" to create a 2200 ELO engine. When I started to write Maverick I also thought it would be trivial to write a strong engine. But having brought Maverick up to about 2300 ELO I would say it isn't "trivial "to write a decent engine. There are many moving parts; all of which need to be implemented. I think a better description is "straightforward". In other words, if you research what is needed to create a strong engine there is clear path. But you still need to put the time in and code (and debug). This isn't trivial - maybe it's just semantics.
I put may be 150 hours on it, and I work in programming for more than 20 years, so can't say it's trivial.
Daniel José - http://www.andscacs.com
-
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Re: New engine - Andscacs.
Arranged the high cpu usage. I put a sleep when is waiting and not pondering. So better discard the old version.
Daniel José - http://www.andscacs.com