IPCCC Paderborn 2007

[swami]

http://wwwcs.uni-paderborn.de/~IPCCC/

IPCCC 07 begins day after tomorrow, only few programs have so far registered but they are all strong.

Code: Select all

1.Diep	
2.Gandalf	
3.Gridchess	
4.Hiarcs	
5.Isichess	
6.Jonny	
7.Rybka	
8.Deep Shredder	
9.Spike	
10.ParSOS

Is Gridchess eligible in all of the tournaments? I thought it heavily borrows code from Fruit and Crafty,parts of crafty/Toga/Fruit with some changes made and runs on a high performance cluster?

anyway,i think Rybka would win this, with Deep Shredder/Hiarcs/Gridchess in for a close fight for the second spot.

[Tony Thomas]

It is disturbing to see Gridchess competing again. I guess it probably runs with the latest Toga engine this time around with added multi processor capabilities.

[Peter Skinner]

It is disturbing to see Gridchess competing again. I guess it probably runs with the latest Toga engine this time around with added multi processor capabilities.

If I were a competitor there, I would put in a _very_ objection to it being allowed to participate.

Gridchess is not open source, but uses open source code from various other engines. Those authors should be stepping up and protecting their sport from such programs.

Imagine if Rybka, Shedder, Hiarcs, or Fritz were released as open source... dear lord... true amateur programmers would have no choice but to clone as well just to keep up.

This is setting a nasty trend. One thing is absolutely certain. I will not be accepting an entry to CCT 10 for Gridchess.

[Gandalf]

Gridchess is not open source, but uses open source code from various other engines. Those authors should be stepping up and protecting their sport from such programs.

As far as I know, as long as they are not distributing their modified version of Toga they aren't violating the GPL.

[swami]

Gridchess is not open source, but uses open source code from various other engines. Those authors should be stepping up and protecting their sport from such programs.

As far as I know, as long as they are not distributing their modified version of Toga they aren't violating the GPL.

But they can't participate in official tournaments such as this and WCCC etc,right?

you mean someone can make some few changes, implement mp support and take it to any official tournament?maybe GPL is incomplete and doesn't state anything about this?

[hgm]

you mean someone can make some few changes, implement mp support and take it to any official tournament?maybe GPL is incomplete and doesn't state anything about this?

This is not a violation of copyrights. It is entirely up to the tournament organizers if they allow such a thing. If you don't want your code to be used this way... Don't make it public!

Furthermore, if the authors of the original code don't participate with the original engines, and back the one who used their code, it is hard for tournament organizers to refuse such an engine. What should they base such a refusal on? "This engine is a cooperative effort, between you and a well known programmer of another engine, so you can't participate"? What about Spike or Dirty, which has two authors that each have written engines before? Should such programs be excluded from tournaments?

[swami]

you mean someone can make some few changes, implement mp support and take it to any official tournament?maybe GPL is incomplete and doesn't state anything about this?

This is not a violation of copyrights. It is entirely up to the tournament organizers if they allow such a thing. If you don't want your code to be used this way... Don't make it public!

Furthermore, if the authors of the original code don't participate with the original engines, and back the one who used their code, it is hard for tournament organizers to refuse such an engine. What should they base such a refusal on? "This engine is a cooperative effort, between you and a well known programmer of another engine, so you can't participate"? What about Spike or Dirty, which has two authors that each have written engines before? Should such programs be excluded from tournaments?

you make a good point, for such an engine(gridchess) this need to obtain a permission from both the authors Bob and Fabien before registering for such an event, Gridchess team also do need some kind of agreement between Fabien and Bob too!

Did Fabien and Bob agree together in the first place?some kind of mutual understanding and agreement/deal about how much of their code is being used in a single engine and how much credit do they deserve each or whether it constitutes team etc?No, Afaik, correct me if I'm wrong because I don't think that Fabien was active around the time that Grid chess took part in the WCCC.Only Bob had so far responded.

[swami]

From the link:

http://www.grappa.univ-lille3.fr/icga/p ... php?id=520

Code: Select all

GridChess is composed of two major components: 1) The proxy chess engine (Crafty based) performs no tree search itself but has some kind of a master role to control the optimistic pondering with distributed worker clients. As a simplified explanation of optimistic pondering here, one can imagine the worker clients forming a pondering pipeline with expected opponent moves extracting this information from the principal variations provided by the chess engines. 2) Real chess engines (controlled by distributed worker clients), Fruit/Toga based, parallelized with Young Brothers Wait Concept (YBWC). This way a combination of two parallel concepts was realized building the complete GridChess system: The parallel Fruit/Toga base engines using the YBWC may run on high performance clusters, each cluster representing a worker client for the proxy chess engine. Several such clusters are then used for an asynchronous distributed game-tree search with the optimistic pondering method.

I think the Gridchess needed 3 main programmers to agree upon themselves, so that means it is a team of 5 programmers counting two more members from Gridchess team, unanimous agreement is needed.Did they all come into such agreement?If yes, then there are more questions like, for example how or why it should be allowed in WCCC ? etc

[hgm]

According to the organizers of WCCC 2007, the GridChess team did show them permission in writing from Bob and Fabien to enter WCCC. Under WCCC rules this would exclude Crafty and Fruit from participating, which was no problem as they had no intention of participating, and indeed did not. In tournaments were an author is allowed multiple entries if they are sufficiently different, even that would not be a an issue, depending on what was defined as 'sufficiently different'.

I suppose the same holds for the Paderborn tournament.

[Daniel Mehrmann]

you mean someone can make some few changes, implement mp support and take it to any official tournament?maybe GPL is incomplete and doesn't state anything about this?

This is not a violation of copyrights. It is entirely up to the tournament organizers if they allow such a thing. If you don't want your code to be used this way... Don't make it public!

Furthermore, if the authors of the original code don't participate with the original engines, and back the one who used their code, it is hard for tournament organizers to refuse such an engine. What should they base such a refusal on? "This engine is a cooperative effort, between you and a well known programmer of another engine, so you can't participate"? What about Spike or Dirty, which has two authors that each have written engines before? Should such programs be excluded from tournaments?

Yes, from the law/license side everything looks fine. But it's well known, at least for the organizers, that a lot of participants of the last years and maybe new authors will not participant as long such engines will be allowed to play in this tournament.

It's a choice of the organizers to allow gridchess and have less participants or disallow it and have more participants (programmers).

However, i wouldn't participant in the case of Gridchess will be in.

Best
Daniel