ethical dilemma

[jdart]

I think that we have at most mix of 2 programs (rybka and fruit)

As I have said in another thread, I think you have to distinguish between use of algorithms and literal copying of code. It appears to me there is very little literal code borrowed from Fruit. Algorithms are not protectable by copyright.
I can't say about borrowing from Rybka, not having original source for that. But my first take is that little of Strelka would be covered under the Fruit copyright.

--Jon

[Uri Blass]

I think that we have at most mix of 2 programs (rybka and fruit)

As I have said in another thread, I think you have to distinguish between use of algorithms and literal copying of code. It appears to me there is very little literal code borrowed from Fruit. Algorithms are not protectable by copyright.
I can't say about borrowing from Rybka, not having original source for that. But my first take is that little of Strelka would be covered under the Fruit copyright.

--Jon

"at most" means that I was not sure that it is even a mix of 2 programs.

My impressions are similiar to you that very little literal code borrowed from fruit but when people insisted that it is not the case including some programmers I felt less sure about my opinion so I wrote "at most".

Uri

[Rolf]

It appears that it was ethically wrong to disassemble Rybka in the first place, but I think it was the best decision to make the source available to all people, instead of making them available to only 'selected' people. As soon as some persons got the source, and could analyze or modify it, I felt that it was only fair if everybody had the chance to do so. So I'm glad the sources are available now. Furthermore, the Rybka version it derives from is 2 years old, as I understand.

I don't really want to get in on this discussion, but I don't really understand this. Publishing the sources from strelka was of course no friendly act. The people that do this, they are just the equivalent of programming hooligans, or whatever term you want to come up with, they do this for the attention they are getting and the interest people have in learning about programming ideas that were not meant to be made public by the author.

Is it okay to rob a bank as long as you don't keep the money for yourself but give it away to everybody else, stealing from the rich and giving to the poor? Osipov as Robin Hood? 'Hood' is right, Robin Hood I don't think so...

There is really no waterproof programming way to protect the intellectual property of programmers ideas for long by encryption, obfuscation or whatever, but if a whole community of looters actively would start banding together to decipher commercial programs, chess programs in this case, publishing the sources for everybody, to spread as many clones as possible, under any name they can come up with, what chance do you stand as a lone commercial programmer against that?

This does not hurt computer chess? Would you justify this? Come on people!

Eelco

OK, then what about the people that come here, ask questions, get lots of ideas and algorithms from active programmers, then they find a new idea, hide it and go commercial. I think they are "hooligans" just as much as this case.

Robert,
if we extend your example then "all those students who go to universities and later invent something and use their invention to become successful are also hulligans." ... that doesn't seem right. The reason is that there is no equivalence, neither in spirit nor intent, that can be drawn between a forum where people exchange ideas, learning from one another, hoping to invent something and the action of stealing the unique ideas that someone might actually have.

Regards,
Hristo

For me, that analogy doesn't work. Here's why. At the university, there is a specific "quid-pro-quo" between faculty and students. Students pay tuition, which pays our salaries. We, in turn, teach the students about various subjects. There is a two-way interchange.

Between many of us here, there is a two-way interchange. We discuss ideas, we exchange ideas, we make suggestions, we might keep secrets for a tournament, but then we reveal what we are doing (in my case, this is pretty obvious since I release source).

The example I cited was missing exactly 1/2 of that. Discuss ideas, ask questions, even get pointers that take you in a good direction, but once you discover something new and different, clam up...

Not what we in academia do at all, which was my point...

Exactly this claim is totally wrong. In short you have it all wrong and I stand for this verdict 100%. Here is why:

1) You yourself speak of a practice you'd adopted from the moment on when you released your open source. And than then you held certain secrets hidden for a while but after a tournament you were always showing up with your latest findings. Well, that is nice to know but it has nothing to do with the real world of competive sports. It does also contradicts the whole history of chess as such. As a player you always keep your best findings also if it might take decades before a relevant opponent serves you the opportunity. As a chessplayer you know all that. So why giving a different picture?

2) But you are also totally wrong for what you like to call academia. Everything that is sponsored by business interested isnt made public. In special if it has military relevance. There you dont just hide the findings but as the IBM "research" with Deep Blue has reveiled, even basics of science are "neutralized" to be able to cheat your own client who should have allowed you to test your achievements. Reason is just that IBM paid it all. We had extended debates about it and you were always defending this "scientists" because they had been standing under contract with IBM. Yes, but I objected, also a contract cant allow you to forget about fundamental rules of science and that every result that you got under such distorted conditions is worthless. The indistry should then just pay a scientist for what he should make up in their interest. Nobody would believe such "results". But you are still today believing that 'psyching out' a chessplayer would prove the class act of a machine chessplayer. Ridiculous. With such a position taken by a most famous computer sciences expert and professor you undermine the ethical basis of a whole field.

3) A case like the one from the last CCT with HIARCS operating Harvey Williamson is such a typical consequence if you once twist the meaning of science rules. If only the outcome interests then the details of the achievement are of secondary interest. But this is exactly what is called cheating in science. Science means that no matter how hard you try under methodical laws of science IF you dont get the wanted result you simply dont get it and you cant get it by fizzling here a bit and there with the conditions of your experimental design. If however this is done in a famous case like the IBM Deep Blue Hsu case, then also Harvey Williamson is entitled to try all he can to make the conditions favorable for HIARCS. No mistake - I know that exactly you have criticised this little scandal, as it was not in accordance with the whole tournament practice that you had experienced and mainly defined yourself as a participant.

4) And now this Strelka scandal. Where did you get from that Vasik Rajlich had published and defined his Rybka Beta code as open source? If he gave the program for free, the situation is still different from the one of FRUIT. But here we have still another problem. There is someone, perhaps a whole group, that threates criminally to publish the code of RYBKA 2.3.2a and you dont comment on that data but you make instead a few condescenting remarks on the lifetime of the RYBKA project. This is when I lose the old Bob out of focus. But to the contrary, you also never condemned anything in connection with RYBKA. But then, if it looks kosher and nothing but kosher, you still want to dismantle this fantastic creation so that Vas loses his advantage over his competition? Why? Because of true sciene? I dont believe you with this academia claim.

[jdart]

I think this thread is discussing two different issues.

Reverse engineering and publishing a closed, commercial source program is not a good thing (IMO) .. basically, you are taking commercial trade secrets and revealing them, clearly to the injury of the person who had them in the original program.

But using an algorithm or idea from a published source, whether it be an academic article, a forum like this, or an open source program, is a different matter, I think, so long as you are not wholesale copying another author's work, or claiming it as your own (Note also that GPL explicitly permits complete copying - as long as you also reveal source when/if you publicly distribute it).

[gerold]

I think this thread is discussing two different issues.

Reverse engineering and publishing a closed, commercial source program is not a good thing (IMO) .. basically, you are taking commercial trade secrets and revealing them, clearly to the injury of the person who had them in the original program.

But using an algorithm or idea from a published source, whether it be an academic article, a forum like this, or an open source program, is a different matter, I think, so long as you are not wholesale copying another author's work, or claiming it as your own (Note also that GPL explicitly permits complete copying - as long as you also reveal source when/if you publicly distribute it).

Hi Jon. TOGA did not reveal source when it was distributed . And
claimed it as his own work. Some think toga is not a clone.
Do you think it is legal computer chess program or clone.

[jdart]

Toga was originally distributed under the GPL (as was Fruit, from which it derives). So in principle any derived work that is also distributed should be open source and also under the GPL. I haven't kept track of Toga but if it's now closed source and also downloadable they are in violation of GPL, it would seem to me. But I think only Fabien, as the original copyright holder, would be in a position to complain about it.

--Jon

[Christopher Conkie]

Toga was originally distributed under the GPL (as was Fruit, from which it derives). So in principle any derived work that is also distributed should be open source and also under the GPL. I haven't kept track of Toga but if it's now closed source and also downloadable they are in violation of GPL, it would seem to me. But I think only Fabien, as the original copyright holder, would be in a position to complain about it.

--Jon

Sorry Jon but Gerold is right. Originally Toga was given to testers under the assumption it was Thomas's own work with no GPL. Only after discovery that it was Fruit (not hard to see btw) did he come clean about it. Any of the testers will tell you this.

Then only after all this was the GPL included along with the source at official release time.

Christopher

[jdart]

Well, GPL is very clear about what is required here. And Thomas is complying now it seems.

Personally, I am not a big GPL fan - I think RMS is a bit of a nut because he thinks this is the One True License and even perfectly fine open source projects like Apache are evil, not to mention closed source, which RMS thinks shouldn't exist. But it has its merits for chess software because there is a lot of concern there about closed, binary redistribution of stuff that started out as open source. And GPL disallows this very explictly.

However, it still doesn't stop someone using algorithms from a free program source w/o literal copying. I think we have to just accept this.

--Jon

[Graham Banks]

Mainly the legally dubious threats regarding disassembling several engines and releasing the source code without permission from the engine authors.

Regards, Graham (moderation).

[bob]

It appears that it was ethically wrong to disassemble Rybka in the first place, but I think it was the best decision to make the source available to all people, instead of making them available to only 'selected' people. As soon as some persons got the source, and could analyze or modify it, I felt that it was only fair if everybody had the chance to do so. So I'm glad the sources are available now. Furthermore, the Rybka version it derives from is 2 years old, as I understand.

I don't really want to get in on this discussion, but I don't really understand this. Publishing the sources from strelka was of course no friendly act. The people that do this, they are just the equivalent of programming hooligans, or whatever term you want to come up with, they do this for the attention they are getting and the interest people have in learning about programming ideas that were not meant to be made public by the author.

Is it okay to rob a bank as long as you don't keep the money for yourself but give it away to everybody else, stealing from the rich and giving to the poor? Osipov as Robin Hood? 'Hood' is right, Robin Hood I don't think so...

There is really no waterproof programming way to protect the intellectual property of programmers ideas for long by encryption, obfuscation or whatever, but if a whole community of looters actively would start banding together to decipher commercial programs, chess programs in this case, publishing the sources for everybody, to spread as many clones as possible, under any name they can come up with, what chance do you stand as a lone commercial programmer against that?

This does not hurt computer chess? Would you justify this? Come on people!

Eelco

OK, then what about the people that come here, ask questions, get lots of ideas and algorithms from active programmers, then they find a new idea, hide it and go commercial. I think they are "hooligans" just as much as this case.

Robert,
if we extend your example then "all those students who go to universities and later invent something and use their invention to become successful are also hulligans." ... that doesn't seem right. The reason is that there is no equivalence, neither in spirit nor intent, that can be drawn between a forum where people exchange ideas, learning from one another, hoping to invent something and the action of stealing the unique ideas that someone might actually have.

Regards,
Hristo

For me, that analogy doesn't work. Here's why. At the university, there is a specific "quid-pro-quo" between faculty and students. Students pay tuition, which pays our salaries. We, in turn, teach the students about various subjects. There is a two-way interchange.

Between many of us here, there is a two-way interchange. We discuss ideas, we exchange ideas, we make suggestions, we might keep secrets for a tournament, but then we reveal what we are doing (in my case, this is pretty obvious since I release source).

The example I cited was missing exactly 1/2 of that. Discuss ideas, ask questions, even get pointers that take you in a good direction, but once you discover something new and different, clam up...

Not what we in academia do at all, which was my point...

Exactly this claim is totally wrong. In short you have it all wrong and I stand for this verdict 100%. Here is why:

1) You yourself speak of a practice you'd adopted from the moment on when you released your open source. And than then you held certain secrets hidden for a while but after a tournament you were always showing up with your latest findings. Well, that is nice to know but it has nothing to do with the real world of competive sports. It does also contradicts the whole history of chess as such. As a player you always keep your best findings also if it might take decades before a relevant opponent serves you the opportunity. As a chessplayer you know all that. So why giving a different picture?

2) But you are also totally wrong for what you like to call academia. Everything that is sponsored by business interested isnt made public. In special if it has military relevance. There you dont just hide the findings but as the IBM "research" with Deep Blue has reveiled, even basics of science are "neutralized" to be able to cheat your own client who should have allowed you to test your achievements. Reason is just that IBM paid it all. We had extended debates about it and you were always defending this "scientists" because they had been standing under contract with IBM. Yes, but I objected, also a contract cant allow you to forget about fundamental rules of science and that every result that you got under such distorted conditions is worthless. The indistry should then just pay a scientist for what he should make up in their interest. Nobody would believe such "results". But you are still today believing that 'psyching out' a chessplayer would prove the class act of a machine chessplayer. Ridiculous. With such a position taken by a most famous computer sciences expert and professor you undermine the ethical basis of a whole field.

3) A case like the one from the last CCT with HIARCS operating Harvey Williamson is such a typical consequence if you once twist the meaning of science rules. If only the outcome interests then the details of the achievement are of secondary interest. But this is exactly what is called cheating in science. Science means that no matter how hard you try under methodical laws of science IF you dont get the wanted result you simply dont get it and you cant get it by fizzling here a bit and there with the conditions of your experimental design. If however this is done in a famous case like the IBM Deep Blue Hsu case, then also Harvey Williamson is entitled to try all he can to make the conditions favorable for HIARCS. No mistake - I know that exactly you have criticised this little scandal, as it was not in accordance with the whole tournament practice that you had experienced and mainly defined yourself as a participant.

4) And now this Strelka scandal. Where did you get from that Vasik Rajlich had published and defined his Rybka Beta code as open source? If he gave the program for free, the situation is still different from the one of FRUIT. But here we have still another problem. There is someone, perhaps a whole group, that threates criminally to publish the code of RYBKA 2.3.2a and you dont comment on that data but you make instead a few condescenting remarks on the lifetime of the RYBKA project. This is when I lose the old Bob out of focus. But to the contrary, you also never condemned anything in connection with RYBKA. But then, if it looks kosher and nothing but kosher, you still want to dismantle this fantastic creation so that Vas loses his advantage over his competition? Why? Because of true sciene? I dont believe you with this academia claim.

All I will say is that (a) Vas used to ask questions, look at other engines (hence the self-proclaimed fruit influence on his program among other things) and so forth. Then when he discovered something new, off he went. Compare that with the _wealth_ of computer chess papers describing everything from iterated deepening, to hash table implementations, to bitboards, to null-move, to singular extensions, to endgame databases, to you-name it. There is a big difference.

(b) I am completely unconcerned about the reverse-engineering that has been done. Seems like a fair way to "even the playing field" by forcing a secretive author to expose secrets he has desparately tried to hide by obfuscation of this PV, depth and node counter displays. I'm not going to lose any sleep over this at all. It isn't my concern...