Sloppy 0.1.1 released

[Tord Romstad]

But chess is a game, namly a strategy board game as many other games too. It's a competition between programmers and her babies.

That depends on your point of view. To me, computer chess is definitely not a competition between programmers, but a big collective effort with the goal of improving the state of the art in the algorithms for strategic game playing. I have no interest whatsoever in the competitive side of computer chess. I like to see my program improve, but I am even happier if other programs improve by the same or a faster rate.

It is possible, perhaps even likely, that your view is more common than mine, but both views are equally valid. I am pretty that Turing and Shannon, the true pioneers of chess programming, were closer to my position.

So, the basic idea behind GPL doesn't work very well here.

To me, it works extremely well. Once again, it depends on your goals in chess programming.

If you ask developers, like Fabien (Fruit/Toga) or Tord (Glaurung), why they are released the versions under the GPL you'll get the answer that they only want that other rookies can learn how to write, or better how it works, a chessprogram and not copy the content in any case.

Not if you ask me. People are welcome, and even encouraged to copy any part they want of my program if they find it useful, as long as they observe the GPL.

Tord

[Ryan Benitez]

But chess is a game, namly a strategy board game as many other games too. It's a competition between programmers and her babies.

That depends on your point of view. To me, computer chess is definitely not a competition between programmers, but a big collective effort with the goal of improving the state of the art in the algorithms for strategic game playing. I have no interest whatsoever in the competitive side of computer chess. I like to see my program improve, but I am even happier if other programs improve by the same or a faster rate.

It is possible, perhaps even likely, that your view is more common than mine, but both views are equally valid. I am pretty that Turing and Shannon, the true pioneers of chess programming, were closer to my position.

So, the basic idea behind GPL doesn't work very well here.

To me, it works extremely well. Once again, it depends on your goals in chess programming.

If you ask developers, like Fabien (Fruit/Toga) or Tord (Glaurung), why they are released the versions under the GPL you'll get the answer that they only want that other rookies can learn how to write, or better how it works, a chessprogram and not copy the content in any case.

Not if you ask me. People are welcome, and even encouraged to copy any part they want of my program if they find it useful, as long as they observe the GPL.

Tord

Tord, I respect you very much as a programmer and for your way of thinking. The problem in computer chess is that because it is a game vs. another opponent is will naturally attract competition. This competition in a perfect world would go just fine with the GPL world if not for the huge flaw in humans. That flaw is that some of them are dishonest and would rather steal from you than observe the GPL. Maybe the saddest part is that they would gain respect observing the GPL far more than cheating it.

Ryan

[Uri Blass]

"big collective effort with the goal of improving the state of the art in the algorithms for strategic game playing"

Based on this logic the next logical step is simply to buy the source code of rybka in order to release it and help improve the state of the art in the algorithms for strategic game playing.

I guess that if you pay vasik enough money he is going to release the source of rybka.

With big collective effort if you find million people who pay 50$ to vasik not in order to buy rybka only for themselves but in order to make the source free then you may get rybka with free source code.

I did not ask Vasik if he is going to release the source of rybka2.3.2a for 50 million dollars but I am almost sure that he is going to be very happy to do it.

The main problem is that I think that you will not find many people who are going to pay him 50$ for that so the collective effort to improve the state of the art is not collective effort but effort of only few people.

Uri

[Thomas Mayer]

Hi Tord,

But chess is a game, namly a strategy board game as many other games too. It's a competition between programmers and her babies.

That depends on your point of view. To me, computer chess is definitely not a competition between programmers, but a big collective effort with the goal of improving the state of the art in the algorithms for strategic game playing. I have no interest whatsoever in the competitive side of computer chess.

well, you took part in computer chess championships as well. So I wonder if you think those competitions would make any sense if they would end up in something like 10 Glaurungs fight against 50 Fruits/Togas + some patchworkengines. Can I get an answer on this ?

Greets, Thomas

[ilari]

"big collective effort with the goal of improving the state of the art in the algorithms for strategic game playing"

Based on this logic the next logical step is simply to buy the source code of rybka in order to release it and help improve the state of the art in the algorithms for strategic game playing.

Actually that wouldn't be completely unheard of. I remember a collective effort to buy and make free a commercial role-playing game that couldn't be released otherwise. This doesn't happen often because most open-source projects are non-profit and the developers don't have the cash or will to do this.

However, there are many cases where the developers of a proprietary program switched to GPL, and thanks to the collective effort of the open source community the program became a lot better, Blender is probably the best example of this.

[Tord Romstad]

Tord, I respect you very much as a programmer and for your way of thinking.

The respect is mutual.

The problem in computer chess is that because it is a game vs. another opponent is will naturally attract competition. This competition in a perfect world would go just fine with the GPL world if not for the huge flaw in humans. That flaw is that some of them are dishonest and would rather steal from you than observe the GPL.

Some of them are dishonest, but the number is very small. I agree with Dann that the perpetual witch-hunts and the general atmosphere of suspicion and jealousy whenever a new strong engine appears is a far bigger annoyance than the clones by themselves.

I also don't agree with the phrase "steal from you". The word "steal" is wrong, because neither I nor anybody else loses anything. "From you" also sounds wrong to my ears, because I don't regard my code as my property. It belongs to the community, not to me.

Breaking the GPL is illegal, but it is not theft.

Maybe the saddest part is that they would gain respect observing the GPL far more than cheating it.

Yes. I have great respect for Ilari and his program.

Tord

[Tord Romstad]

well, you took part in computer chess championships as well.

I did (in Mainz in 2005 and 2006, and in Lodz this year). It's a great way to meet people and make new friends.

So I wonder if you think those competitions would make any sense if they would end up in something like 10 Glaurungs fight against 50 Fruits/Togas + some patchworkengines. Can I get an answer on this ?

Most tournaments don't allow clones or derived works to participate. I am fairly sure the majority of programmers and other computer chess enthusiasts (including myself) would find a tournament without this rule to be less interesting, it therefore seems likely that most future computer chess tournaments will keep using the no-clones rule.

Perhaps there are also some people who would like to see or play in tournaments which are open to all sorts of clones, and I wouldn't complain if they started organizing such tournaments, or if some of the participants used my engine (with or without additions of their own). I probably wouldn't be very tempted to participate in such a tournament myself, though.

Tord

[Christopher Conkie]

well, you took part in computer chess championships as well.

I did (in Mainz in 2005 and 2006, and in Lodz this year). It's a great way to meet people and make new friends.

So I wonder if you think those competitions would make any sense if they would end up in something like 10 Glaurungs fight against 50 Fruits/Togas + some patchworkengines. Can I get an answer on this ?

Most tournaments don't allow clones or derived works to participate. I am fairly sure the majority of programmers and other computer chess enthusiasts (including myself) would find a tournament without this rule to be less interesting, it therefore seems likely that most future computer chess tournaments will keep using the no-clones rule.

At last, that's what im looking for. In the case of Sloppy there is no witch hunt. It is obvious to a blind man that the weights are Toga's and it is said that is so by the author (of the rest of the engine) himself. Maybe he will change them, find his own....maybe better ones and then be able to enter those tournaments which are indeed very interesting and require original works. I hope so anyway.

Perhaps there are also some people who would like to see or play in tournaments which are open to all sorts of clones, and I wouldn't complain if they started organizing such tournaments, or if some of the participants used my engine (with or without additions of their own). I probably wouldn't be very tempted to participate in such a tournament myself, though.

Tord

I also would not have a problem with that either. Provided it is stated that ithose rules applied no one could complain about anything like that. I do however understand that the challenge against different engines is much more interesting than one where many of the engines are extremely similar.

I just dont want to see a WCCC or a Leiden full of Fruit's or Toga's or even Glaurung's.

Christopher

[Dann Corbit]

He forgot to mention C who gets something better than the rest and is commercial.

I forgot it because it's not relevant.

I would not bet on that if i were you.

http://64.68.157.89/forum/viewtopic.php ... 28&t=17305

Where do you draw the line? Personally I think it should be 1800 ELO.

I draw the line at copying the evaluation weights exactly of another program.

That's because you do not understand what GPL means.

[Dann Corbit]

Hi Tord,

But chess is a game, namly a strategy board game as many other games too. It's a competition between programmers and her babies.

That depends on your point of view. To me, computer chess is definitely not a competition between programmers, but a big collective effort with the goal of improving the state of the art in the algorithms for strategic game playing. I have no interest whatsoever in the competitive side of computer chess.

well, you took part in computer chess championships as well. So I wonder if you think those competitions would make any sense if they would end up in something like 10 Glaurungs fight against 50 Fruits/Togas + some patchworkengines. Can I get an answer on this ?

Greets, Thomas

This notion that everyone must write a computer chess engine completely from scratch is both a fiction and a perversion.

It is a complete fiction because:
Every chess engine on earth borrows heavily from what others have done. If not from actual code, then definitely from papers and books.

It is an utter perversion because:
It discourages the exchange of information and the advance of the science.

There was a big complaint about piece square tables matching recently.
I examined 26 piece square tables, chosen at random. Most of them have 'pcsq' in their name. Over 70% have a value match with another program. (Keep in mind that piece square tables have been published in chess papers and even offered freely for use in this very newsgroup).

As for a chess competition, here is what I would like to see:
The best ideas from Glaurung and Scorpio and Toga and every other chess program should be considered a baseline. Now, I do not approve of cut and paste. But I think that if you are a chess programmer and you do not examine the other open source chess programs and learn from them, that is a defect in the programmer. Come now, in other areas of computer science we do that and it is a virtue, saving money for our clients and teaching us to be better programmers.

So what about the programmer who does not want to learn and advance by reading other programs, etc.? I think he has chosen his own disadvantage.

Personally, I would like to see every chess programmer at a contest exchange disks that contain their source code with each of the other programmers. Now, I do realize that some programs are commercial and so my suggestion is totally impractical. But in an ideal world, that would be my choice.

It seems to me that the competition aspect has gone too far, because it seems to be very discouraging against sharing information. That (to me) is a clear and strong negative.

I do realize that I am probably in the minority on this and also that my opinion is no better than anyone else's.