Winboard Forum

Greetings every engine coder/programmer,

If anyone of you want to have your engine included in the Hammer of Thor
rating list, please send me an email at mordor(at)halden.net . It can be
beta versions of course, but that particular version will stay in the list
for some time to get a proper amount of games.

I would also like to test some private engines. I usually contribute
to the CCRL rating list, but now these days I am running even
longer time controls in my own little rating list.

Hammer of Thor is a tournament time-controlled rating list (40/120-20/60-+30).

I am also a member of the CCC for a while, check my similar thread
at the CCC. http://216.25.93.108/forum/viewtopic.php?t=13054

Best Regards,
Martin

If engines use their own book I am interested. Please describe your tournament conditions and rules.

Ron

Hello Ron,

Sorry but I use a common book for all engines,
because I'm used to it from the CCRL rating list.

Here are the two PC's used for the rating list:

Zap!Chess 64-bit GUI
AMD Athlon X2 Dual Core, 2568 MHz
256 MB Hash (2CPU)
128 MB Hash (1CPU)
HS-8moves.ctg
3-4-5 Piece Tablebases
(3-4-5 Piece Bitbases)

Time control is 40 moves in 120 minutes, then 20 moves in 60 minutes, then 30 minutes for the rest.

Fritz 10 GUI
AMD Athlon 64, 2210 MHz
128 MB Hash (1CPU)
HS-8moves.ctg
3-4-5 Piece Tablebases
(3-4-5 Piece Bitbases)

Time control is 40 moves in 136 minutes, then 20 moves in 68 minutes, then 34 minutes for the rest (Adapted to 40/120).

Anyway, thank you for your interest!

Just curious how you implement this time control. The Winboard protocol does not seem to provide in a command to tell the engine how many moves it still has to do in the available time as specified by the 'time' command. So if initially you set it for 40/120, after you award it the next 60 minutes, how does the engine know that it has to play 20 moves in the next 60 minutes, and not 40?

And especially the 30 minutes for the rest of the game. Are you sending it a new 'level' command during the game? The Winboard protocol does not specify how such a command should be interpreted. My engine would simply think that he would have to do another 20 moves in those 30 minutes, and would use almost all time left on move number 20...

All winboard engines are run through the Wb2Uci adapter.

None of the winboard engines I currently run has any problems with the time management.

H.G.Muller wrote:Just curious how you implement this time control. The Winboard protocol does not seem to provide in a command to tell the engine how many moves it still has to do in the available time as specified by the 'time' command. So if initially you set it for 40/120, after you award it the next 60 minutes, how does the engine know that it has to play 20 moves in the next 60 minutes, and not 40?

You have to keep track of it in your engine. The Winboard protocol only supports x moves in y minutes. So if you start with 40 moves in 60 minutes, the next time control after 40 moves will always add another 60 minutes and another 40 moves. Winboard protocol does not support multiple time controlls.

And especially the 30 minutes for the rest of the game. Are you sending it a new 'level' command during the game? The Winboard protocol does not specify how such a command should be interpreted. My engine would simply think that he would have to do another 20 moves in those 30 minutes, and would use almost all time left on move number 20...

No new level commands are sent during the game.

I am also a member of the CCC for a while, check my similar thread
at the CCC. http://216.25.93.108/forum/viewtopic.php?t=13054

Best Regards,
Martin

Just to be sure, i don't allow to use Homer in any of your tournaments.

Best,
Daniel

Daniel,

Why is that?

Martin Thoresen wrote:Some moderator please just close this thread.

What is going on here? If you want to contact us please use the email form. In general we don't accept commands especially if they are posted in the forum.

BTW: I really don't understand why your question causes such reactions.

Volker

Volker Pittlik wrote:

Martin Thoresen wrote:Some moderator please just close this thread.

...

Did you edit your post while I reply? the quoted text and the text visible is different!

Volker Pittlik wrote:

Volker Pittlik wrote:

Martin Thoresen wrote:Some moderator please just close this thread.

...

Did you edit your post while I reply? the quoted text and the text visible is different!

Sorry Volker,

I edited the post after my initital disappointment by the general negativity here. I made this thread with a honest intention.

I'll just browse WBEC and use the commercially available engines in their updated list.

Feel free to close this thread if you like.

Thanks,
Martin

Martin Thoresen wrote:...
Feel free to close this thread if you like.

Thanks,
Martin

To be honest: I do not understand the commotion. I see no reason to delete it.

Volker

Daniel Mehrmann wrote:Just to be sure, i don't allow to use Homer in any of your tournaments.

Best,
Daniel

Daniel, why is that?

Martin Thoresen wrote:

Daniel Mehrmann wrote:Just to be sure, i don't allow to use Homer in any of your tournaments.

Best,
Daniel

Daniel, why is that?

Hi Martin !

I was trying to make some pressure.
The people should think about why non WB/UCI engines with own GUI not going to be tested.
All i wanna see was a think progress. Seems it doesn't work.

Ok, you can use Homer again.

Best,
Daniel

Daniel Mehrmann wrote:Hi Martin !

I was trying to make some pressure.
The people should think about why non WB/UCI engines with own GUI not going to be tested.
All i wanna see was a think progress. Seems it doesn't work.

Ok, you can use Homer again.

Best,
Daniel

Hello Daniel and thank you for your reply.

I got several reasons for why I can't test non CB-native/WB/UCI
engines. The biggest reason is because I am a normal person
with a job, a family and a dog. So my time is very limited, that's why
I am dependant of automatic tournaments that can run without
needing my attention/time. I am sure you understand this.

I have tested Homer for CCRL several times, and I really enjoy your
engine. So that's why I was a bit surprised of this sudden veil of
negativity here. If there's a thing I want to avoid is to create some
sort of negativity towards the engine creators. I really envy your
work and more than once I've wished that I knew how to program.

Best Regards,
Martin

Martin Thoresen wrote:

Daniel Mehrmann wrote:Hi Martin !

I was trying to make some pressure.
The people should think about why non WB/UCI engines with own GUI not going to be tested.
All i wanna see was a think progress. Seems it doesn't work.

Ok, you can use Homer again.

Best,
Daniel

Hello Daniel and thank you for your reply.

I got several reasons for why I can't test non CB-native/WB/UCI
engines. The biggest reason is because I am a normal person
with a job, a family and a dog. So my time is very limited, that's why
I am dependant of automatic tournaments that can run without
needing my attention/time. I am sure you understand this.

I have tested Homer for CCRL several times, and I really enjoy your
engine. So that's why I was a bit surprised of this sudden veil of
negativity here. If there's a thing I want to avoid is to create some
sort of negativity towards the engine creators. I really envy your
work and more than once I've wished that I knew how to program.

Best Regards,
Martin

Well, it was no personal attack against you. You maybe had just bad luck, because you was on the wrong place at the wrong time. It could hit each of the testers of both groups.

What i want to say is for both groups in general:

http://www.talkchess.com/forum/viewtopi ... 21&t=13054

Once more, i understand manual testing costs a lot of time. I understand that not so much games are possible. I understand personal time problems, like Martin, as well.

I don't want to force big testing of GUI based engines now without protocol support. I want a reaction , i want a think process, i want see a talk on public boards, right now. Thats all what i want. But i see no answer right now, why ?

Well, if one tester maybe testing at the time monday-friday 300 games automaticly in a week, i believe it costs not much testingh 5 games manual on such engines without protocol support.

If each tester only would do this little number of games, maybe in a week, maybe in a month, if the tester has time, we could slightly integrate those engine in the lists !

5 games , max 3 hours in a week or month, is it to much ??
How many hours do you watch your running games ?

Come on guys.

Best,
Daniel

Daniel Mehrmann wrote:Come on guys.
Daniel

I come Daniel !

I think that your series of posts here and on CCC on this topic is a real shame!

You just should be banned of most testing schemes.

The way you attacked this guy personally and the whole corporation of testers is not acceptable in my eyes.

Who are you with your little piece of software to go with such statements?

These people (the testers) give hours and days, freely, without any public recognition for the promotion of computer chess and for the improvement of your program among others.

You should begin with acknowledgements instead of accusations !

If you wish manual testing either do begin with being enough polite and convincing so as to find the kind of guys who will accept to go for this incredibly painful task.

Or do it yourself and stop with your silly comments!

Marc

Marc Lacrosse wrote:

Daniel Mehrmann wrote:Come on guys.
Daniel

I come Daniel !

I think that your series of posts here and on CCC on this topic is a real shame!

You just should be banned of most testing schemes.

The way you attacked this guy personally and the whole corporation of testers is not acceptable in my eyes.

Who are you with your little piece of software to go with such statements?

These people (the testers) give hours and days, freely, without any public recognition for the promotion of computer chess and for the improvement of your program among others.

You should begin with acknowledgements instead of accusations !

If you wish manual testing either do begin with being enough polite and convincing so as to find the kind of guys who will accept to go for this incredibly painful task.

Or do it yourself and stop with your silly comments!

Marc

Hi Marc !

I don't understand you or maybe my english is bad in some cases here. I just pointed out i want a discuss about this topic and i made a example suggestion, but nothing more.

Tester doing a great job, no question. But on the other side i see a lot of programmers which has no choice coming in these lists with her engines.

Forbidden to use Homer was only a try that the people should start thinking about it. It was never a personal attack.

Also if the groups decide doing nothing here with these engines it's ok. I tryed my best to support such programmers. That was all what i want.

Best,
Daniel

Daniel Mehrmann wrote:

Marc Lacrosse wrote:

Daniel Mehrmann wrote:Come on guys.
Daniel

I come Daniel !

I think that your series of posts here and on CCC on this topic is a real shame!

You just should be banned of most testing schemes.

The way you attacked this guy personally and the whole corporation of testers is not acceptable in my eyes.

Who are you with your little piece of software to go with such statements?

These people (the testers) give hours and days, freely, without any public recognition for the promotion of computer chess and for the improvement of your program among others.

You should begin with acknowledgements instead of accusations !

If you wish manual testing either do begin with being enough polite and convincing so as to find the kind of guys who will accept to go for this incredibly painful task.

Or do it yourself and stop with your silly comments!

Marc

Hi Marc !

I don't understand you or maybe my english is bad in some cases here. I just pointed out i want a discuss about this topic and i made a example suggestion, but nothing more.

Tester doing a great job, no question. But on the other side i see a lot of programmers which has no choice coming in these lists with her engines.

Forbidden to use Homer was only a try that the people should start thinking about it. It was never a personal attack.

Also if the groups decide doing nothing here with these engines it's ok. I tryed my best to support such programmers. That was all what i want.

Best,
Daniel

What is wrong with being on a rating list. Yes, the programmer does have a choice, he can keep the engine private. If you (the programmer) want to be on the top you (he/she) have to work hard, and also you (he/she) have to be a good programmer. Then again for most people chess is a hobby, have some sportsman spirit daniel. Dont attack a rating list because your engine is doing bad, your engine seem to perform bad under blitz conditions. However, under Leo's conditions your engine is quite a charmer, it made it up to the second division.

Tony Thomas wrote:
What is wrong with being on a rating list. Yes, the programmer does have a choice, he can keep the engine private. If you (the programmer) want to be on the top you (he/she) have to work hard, and also you (he/she) have to be a good programmer. Then again for most people chess is a hobby, have some sportsman spirit daniel. Dont attack a rating list because your engine is doing bad, your engine seem to perform bad under blitz conditions. However, under Leo's conditions your engine is quite a charmer, it made it up to the second division.

eh ?

I'm not talking about my engine. My engine supports UCI and WB2 protocols.

Best,
Daniel