World Computer Chess Championship ?

[Houdini]

1) There is simply no such thing like "... have effectively become public domain". Why do you believe that the original author's rights have "magically" vanished?

These source codes legally ARE public domain.
Whatever you say about the "original author" and his or her "rights" is your personal interpretation of what appears to be a complex case about which neither of us have a lot of information.

2) From the four top engines you have named there is only one that is massively based on the code of "these sources" (you know which one) while the others have reused some of the ideas and concepts within their fully original implementation. It is quite telling that you pretend not to know the difference.

1) What is the relevance of Critter's long journey from Pascal to C, from 8-bit board representation to bitboards, from original to more Ippolitish, when IN THE END the result is nearly indistinguishable from what you call "massively based on code". Why do you attach more value to the journey than to the end result?

2) On the difference between "using code" and "using ideas and concepts".
Your understanding of this issue appears to be rather theoretical and dogmatic.
Marco's comment above shows a lot more insight, I quote Marco:

What it means to steal code ? What it means "I have only taken ideas not code" given that copy & paste is in 99,9% of cases technically impossible and that "to take ideas" in many cases it means to scan at the microscope level the grabbed sources looking for the minimal difference and quickly (monkey) testing anything seemingly different (and finally asking for hints to the forum when the test result is not what was expected).

Personally I object more to someone stealing my ideas than to someone stealing my code. For any good developer the difference between the two is nearly a technical detail. If you give me a good idea, I don't care about your implementation because I will make a better one.
The true value of chess engines lies in the ideas, not in the code: if you steal my ideas, you steal my creative work. If you steal my code without understanding and improving the idea, you're just plain stupid (see Strelka 5...).

Robert

[Rebel]

I would be interested in participating in a tournament similar to the excellent TCEC organized by Martin Thoresen, which had a very well-thought format:

I have to second this. I have no interest in partecipating in tournaments (and this surely will not change in the future), but I really enjoyed Martin's one ! At the fun level it was really much more awesome than the traditional ones. The fact that it didn't pretend to attribute void "titles" to the winner, but just fun to the people watching was a big plus !

Marco, this is all true. However every sport that takes itself seriously has a world championship. We once had an organization that did a good job but then ICGA recently decided to commit suicide, not realizing the times they are living in, not willing to change, and in the hypothetical case they finally would see the light there is this Rybka verdict that will be a millstone around their neck to make any change.

So we are on our own.

What if:

A new organization that organizes a yearly online WCCC. Webcam obliged to avoid cheating, life coverage of every move plus depth, scores and mainlines. Participants have to undergo the similarity tester to be accepted.

Would you play?

Robert ?

Don ?

Richard ?

[Carotino]

I would be interested in participating in a tournament similar to the excellent TCEC organized by Martin Thoresen, which had a very well-thought format:

I have to second this. I have no interest in partecipating in tournaments (and this surely will not change in the future), but I really enjoyed Martin's one ! At the fun level it was really much more awesome than the traditional ones. The fact that it didn't pretend to attribute void "titles" to the winner, but just fun to the people watching was a big plus !

Marco, this is all true. However every sport that takes itself seriously has a world championship. We once had an organization that did a good job but then ICGA recently decided to commit suicide, not realizing the times they are living in, not willing to change, and in the hypothetical case they finally would see the light there is this Rybka verdict that will be a millstone around their neck to make any change.

So we are on our own.

What if:

A new organization that organizes a yearly online WCCC. Webcam obliged to avoid cheating, life coverage of every move plus depth, scores and mainlines. Participants have to undergo the similarity tester to be accepted.

Would you play?

Fantastic!
The Martin's tournament is nice... Even Graham's tournament is funny, but I'd prefer a slightly different formula, with the direct intervention of the programmer (or his agent). With his openings book, where he can choose the variants you prefer, and using their own hardware...
I believe that this kind of tournaments are the future.... I'm sorry, THE PRESENT.

[Carotino]

Last month, I organized a small tournament like this, with the sole purpose of having fun. Four participants, where each one 'maneuver' a chess engine, from his home. It was possible to follow the games in real time...
Well, with very little advertising (almost nothing), every night there were over 5000 spectators. I have also been approached by several sponsors... Of course, I had no serious plan, everything was done for an evening of fun, so I had to refuse several interesting proposals.
I believe that a small fee, and without too much effort, you can organize various tournaments of this type. I also believe that they have a large and enthusiastic audience of chess fans.
Meet in person is a good thing, but in these times of economic crisis, I think many want to avoid covering 10000 kilometers to do so! There is internet or Skype for this.

[Rebel]

Yes, membership, if needed a small yearly fee, no fee for entering, a constitution, trophies, a website, all of it. Two world titles to give away, 2 tournaments, one for equal and one for unlimited hardware. Just some random thoughts.

[Rebel]

These source codes legally ARE public domain.

They are challenged:

http://rybkaforum.net/cgi-bin/rybkaforu ... #pid190922

You are just lucky you are not challenged and Houdini can exist by the grace (or lack of interest) of Vasik.

[MM]

Of course we would never have a ''real'' world champion, but at least, we would have a 100% original world champion.

I don't get this. The Tour de France is not a 'real' Tour de France because doped cyclists are not allowed to participate, and we all know they are way faster than those that ride clean. The Olympics are not 'real' Olympics because doped athletes are banned, and we all know Ben Johnson ran faster than anyone else...

What kind of logic is that?

If clones or derivates are allowed to to partecipate to the chess world championship we would have a ''real'' world champion, where for ''real'' is meant that the world champion is really the strongest engine in the world (clone or not).

That sounds good on paper but the reasoning is badly flawed. Suppose Robert Houdart has 20 different development versions but he doesn't know which version is best? Unless all 20 are allowed to compete then how can we know which engine really is the strongest one? See how silly that is? But based on your (flawed) reasoning that would be perfectly equitable.

Let's assume that most every program is (roughly) about even in strength. Even though that is not the case the principle will be the same. If there are 10 programs then each program would have 1/10 chance of winning the tournament. If you were the only one allowed to bring 2 versions of your program, you would have twice the winning chances of your competitors. That is pretty obviously unfair. So what happens when Ivanhoe, Ippolitto and Robbolito and Firebird are 4 of those 10 programs? The chances that ONE of those 4 will win is actually much higher than ONE program 100 ELO stronger winning.

If we started to play this game then I would have to argue that I should be allowed to not just bring Komodo but several versions of Komodo so that we can "fairly" see which is really the stronger program. If that were not allowed I could just give these various development copies to my friends, they could rename them to something else and even admit that there were branched off of Komodo (but they are different which would be true) and the tournament could not argue with this reasoning or I would just repeat what you said and accuse them of not being interested in seeing which version is really strongest.

So your argument is emotionally appealing but it doesn't really stand up to logic.

This has nothing to do with what is right or wrong, what is legal or not, what is ethical or not.

Take an engine, copy almost all of it, change it in order to make it 50 elo stronger than original...someone could say that is right, legal, ethical but anyway its programmer has a huge advantage to the other 100% original engines' programmers.

In The Tour de France sometimes it's impossible to know who is doped and who's not because everybody knows that doping goes faster than antidoping and today we have cyclers of the past, never banned, who admit to have used doping in the past. Someone is accused to be doped and perhaps he's not. Someone else is not accused of anything and perhaps he's doped.

The same for any other sport.

The question is not ''real or not real'', because, every time that some of all competitors don't partecipate, it's hard to say ''real''.

The question is: how do i know if an engine or an athlete is ok?'' and ''How do i set the rules to allow or not allow an engine/athlete to the competions?''

Hi Don,

I think we are talking about different things or perhaps the same thing, from different points of view.

What i stated is pretty obvious, more obvious if you note that Junior and Hiarcs are the world champions...with Komodo, with Houdini, with Rybka, with Critter ect. ect.

I didn't say what you stated as examples to mean that my reasoning was flawed.

Anyway if Fire, IvanHoe, Igorrit and Robbolito go to the world championship where's the problem? They globally have the biggest chances to win? Yes, and then?

The question is ''which engine is the strongest'', not how many clones or derivates are allowed to partecipate.

Yes, i could give my betas or versions to several friends and they go to the world championship with them. And so? Wouldn't it be a real world championship cause that?

Note that i'm not saying that it should be right...i just say that i expect, from a world championship, that the winner is the strongest engine and if not all engines are allowed i cannot say ''that is the strongest''.

With ''all engines'' of course i don't mean all the versions of a single engine or a lot of clones of the same engine with just a few changes.

That was not what i meant.

That's why i asked the question ''how do i set the rules to allow or not allow an engine to partecipate'', like saying, one should follow a solid and right criterion.

Note that i don't say that playing 20 different versions of Houdini or Komodo in the world championship is right. It's not right and not accettable because the world champiomship is a competition and going to a competition with several version of the same engine is like going to a war with 2 millions soldiers against 1 thousand ones.

But even if it should happen, i say that the winner of this world championship should be the strongest one (''should'' because the games played would be few).

P.S. Hope Komodo 5 soon. Thanks

Best Regards

[Carotino]

Yes, membership, if needed a small yearly fee, no fee for entering, a constitution, trophies, a website, all of it. Two world titles to give away, 2 tournaments, one for equal and one for unlimited hardware. Just some random thoughts.

You can do, just wanting it! I think also it would be something attractive for potential sponsors. But also a golden opportunity for new and young programmers who want to test their work. They should not be the sons of Bill Gates to do this!
Although, of course, the classic tournaments have not lost their charm ...

[Laskos]

Rybka3 - I never looked into the binary. I have R3 equivalent source code from Yuri O.
Rybka4 - Yes, but I did not disclose anything specific besides the fact that it has much lighter eval than R3 (which is much more similar to Ippolit than R3 eval was)

Is it possible that Vas took from Ippos for his Rybka 4?

Kai

[Houdini]

These source codes legally ARE public domain.

They are challenged:

http://rybkaforum.net/cgi-bin/rybkaforu ... #pid190922

You are just lucky you are not challenged and Houdini can exist by the grace (or lack of interest) of Vasik.

Please stop this utter nonsense. What you link to is no legal challenge. Even if it were, the impact for Houdini is nothing, nada, nill.
In what kind of fantasy world do you live if you believe that "Houdini can exist by the grace of Vasik"?

Robert