On escaping copyright by rewriting everything

[Zach Wegner]

Can we know for certain that the borrowing Vas did was not the same as ours?

Yes, depending on who you mean by "ours" of course. Now, this doesn't necessarily implicate code copying. That's why I tend to shy away from the legal aspect of this debate. There are things which I would define as "plagiarism", but might not carry any legal weight. There are situations where the evidence wouldn't cause the FSF to mount a lawsuit, but it would convince others that a wrongdoing has taken place. What are we to do then?

Of course, this does seem like a very provocative statement, but it is not based on opinion. It is about something far more damning and far more concrete than a parser.

[Dann Corbit]

Is every idea in Chess Tiger original?

Some are and some are not.

I guess that every chess program on earth that is worth anything is the same.

I'm not sure why you ask about ideas, as it is clearly established that ideas are not protectable anyway.

I only wanted to establish that all chess programmers borrow ideas from other sources such as articles, books, other programs, conversations, etc.
Can we know for certain that the borrowing Vas did was not the same as ours? In other words, we all know conceptual algorithms are only protected by patent. So the question arises, how were the ideas transformed and is the new work an entity unto itself? Statements about stealing and plagiarism are judgement calls and people who are emotionally involved may make judgements that are colored by their emotions. For that reason, I am calling for restraint. Now, I do not have any problems with investigation of similarities between programs. I even have my own suspicions about certain programs, but because evidence is lacking, I will not make any charges of blame or wrongdoing.

To restate, I have no problem with this sort of thing:
"Look, these two structures are similar"
But I do have a problem with this sort of thing:
"The similarity of these structures indicates that programmer X is a liar and a cheater."

That's all.

// Christophe

I'm prepared to read anything now. I'm pretty sure someone is preparing to declare the copyright laws unconstitutional. Just because they can't apply to Vas.

// Christophe

Of course the laws apply to Vas.
But here is the question:
Is his use of outside information different from yours?
Before you answer, consider:
Rybka is a bitboard program. It is simply not possible to lift any given large fraction and insert it into another program because the internals are different. So in order to use it he would have to understand it and modify it extensively for it to work at all. Is that different than what you did with the program parts you learned from?

Given the existing evidence it is simply not clear that Vas has committed wrongdoing. Or (conversely) if he has then probably all of us are guilty, you included. I am not a lawyer and so I can certainly misunderstand the intent and meaning of the law.

To me, this is clear:
Concepts and ideas are not something that you can own[*]. Code, like a story is copyrightable. That means that your implementation belongs to you. Someone cannot take it and put their name on it and claim that it belongs to them. Similarly, they can't take it, tweak it a bit and claim that it belongs to them. But there is nothing wrong with learning how something works and then using those ideas. Because of studying someone else's work, the new ideas may look very similar. But we know that they simply must be quite different in many places because the underlying plumbing is different. There is also the evidence that it is a heck of a lot stronger, so it is not just a rip-off and tweak. Now, having said all that, it is certainly possible that Vas has done something wrong. But I do not think we should convict him in the court of public opinion. Perhaps being from the U.S. I am used to the idea of people getting due process before we convict them of something.

[*] Unless you have a patent. Distasteful to me, but a fact of life.
[/b]

[Dann Corbit]

Can we know for certain that the borrowing Vas did was not the same as ours?

Yes, depending on who you mean by "ours" of course. Now, this doesn't necessarily implicate code copying. That's why I tend to shy away from the legal aspect of this debate. There are things which I would define as "plagiarism", but might not carry any legal weight. There are situations where the evidence wouldn't cause the FSF to mount a lawsuit, but it would convince others that a wrongdoing has taken place. What are we to do then?

Of course, this does seem like a very provocative statement, but it is not based on opinion. It is about something far more damning and far more concrete than a parser.

Plagiarism cannot be applied in this case (though other crimes could apply). Plagiarism requires failure to give credit where credit is due.

plagiarize, UK USUALLY plagiarise Show phonetics
verb
to use another person's idea or a part of their work and pretend that it is your own:
The book contains numerous plagiarized passages.
If you compare the two books side by side, it is clear that the author of the second has plagiarized (from the first).

Vas gave credit to the following:

"Special Thanks

I hesitate to include this section because I know I'll forget people who have been helpful in this project, but (with advance apologies to the omitted) here goes:

Robert Hyatt - For Crafty. There is nothing like an open source program for passing knowledge to the next generation.
Fabien Letouzey - For Fruit, which shattered a number of computer chess myths, demonstrated several interesting ideas, and made even the densest of us aware of fail-low pruning.
Tord Romstad - For making Fabien aware of fail-low pruning , and more seriously for sharing in every way possible his considerable knowledge.
Eugene Nalimov - For his cryptic but somehow fully functional endgame tablebase access code.
Uri Blass, Gerd Isenberg, Dieter Burssner, Vincent Diepeveen, Raschid Chan, Anthony Cozzie, Mridul M* , Thomas Gaksch, Peter Berger, Sandro Necchi, Ed Shroeder, Amir Ban, Christophe Theron and every one else, past and present, on the computer chess club: For sharing their computer chess knowledge despite the fact that in principle computer chess is a competitive field.
Heinz van Kempen, Guenther Simon, Olivier Deville, Sergio Martinez, Claude Dubois: for testing early versions of Rybka despite countless bugs and annoying problems.
Alex Dumov, Gabriel Luca: for helping a Windows newbie get up to around half-speed without excessive derision (or at least open derision )
and Iweta: for being great! and a pretty good Rybka tester and web master to boot"

[Zach Wegner]

"Rybka is of course original"

--Vasik Rajlich

[Dann Corbit]

"Rybka is of course original"

--Vasik Rajlich

His claims of original ideas does not negate his giving proper credit to those whose ideas inspired him.

The fact that Rybka is 200 Elo better than most top engines is in itself some proof of originality. If it were not novel, everyone would be doing it.

I wrote an external sort routine that uses a priority queue to merge the subfiles. In so doing, it sorts a file of any size in a total of two read and write passes over the data of the file.

I did not invent priority queues. I did not invent merging. Yet I claim that the idea for this sort is original and my invention. A standard polyphase merge takes log2(subfile_count+1) passes over the data, and so it can be considerably slower. In the case of terabyte files, my method can be hundreds of times faster.

[Zach Wegner]

1. We're not talking about ideas
2. We're talking about Rybka 1, not later versions

[Dann Corbit]

1. We're not talking about ideas

Original ideas is a major thrust of what I am talking about. Plagiarism means to claim that you invented something when someone else did. Copyright violation means to go beyond fair use in someone else's publication without permission. If we do not talk about ideas I do not see how we can reach any sort of understanding.

2. We're talking about Rybka 1, not later versions

In that case, 200 Elo would be an exaggeration. Rybka is 100 Elo stronger than Fruit 2.1
http://www.husvankempen.de/nunn/40_40%2 ... liste.html
no Program Elo + - Games Score Av.Op. Draws
123 Rybka 1.0 Beta w32 2815 7 7 6453 61.3% 2736 33.7%
199 Fruit 2.1 2714 10 10 3298 57.7% 2660 34.7%

[kranium]

The fact that Rybka is 200 Elo better than most top engines is in itself some proof of originality. If it were not novel, everyone would be doing it.

Dann-

No one is arguing that Rybka 1.0 is 100% identical to Fruit 2.1. There are many differences, and an improvement in strength is definitely one of them.
The questions and concerns being raised are about the origins of Rybka 1.0, and these (questions and concerns) are perfectly logcial:

how would you (or anyone) answer this question:

if Stelka is known to be a clone of Rybka, (stated as such in Wikipedia, and confirmed by Vas himself), and Strelka is known to have so much Fruit code in it...
isn't it fair and correct to question the legitimacy of both programs??

PS: we now have Strelka 2.0 source code...clearly containing many hundreds of lines of source code taken directly from Fruit 2.1, yet Strelka is still being distributed, and is even being tested by the CCRL and the CEGT? It has been accepted by the chess community?

Norm

[Graham Banks]

we now have Strelka 2.0 source code...clearly containing many hundreds of lines of source code taken directly from Fruit 2.1, yet Strelka is still being distributed, and is even being tested by the CCRL and the CEGT? It has been accepted by the chess community?

Norm

CCRL hasn't tested Strelka for a long while and I don't think that CEGT has tested it at all.
It has been accepted by some, but not by others.

[kranium]

we now have Strelka 2.0 source code...clearly containing many hundreds of lines of source code taken directly from Fruit 2.1, yet Strelka is still being distributed, and is even being tested by the CCRL and the CEGT? It has been accepted by the chess community?

Norm

CCRL hasn't tested Strelka for a long while and I don't think that CEGT has tested it at all.
It has been accepted by some, but not by others.

Strelka 2.0 B 32-bit in CCRL 40/4, 2008-08-25
Ponder off, General books (up to 12 moves), 3-4-5 piece EGTB
Time control: Equivalent to 40 moves in 4 minutes on Athlon 64 X2 4600+ (2.4 GHz)
Tested by CCRL team, 2005-2008, http://computerchess.org.uk/ccrl/404/

CCRL 40/4 main list:
Strelka 2.0 B 32-bit has no rank with rating of 2978 ELO points (+43 -41),
based on 248 games: 166 wins, 26 losses and 56 draws
Score: 78.2%, Average opponent: −215.3, Draws: 22.6%


Pairwise results:
Opponent ELO Score LOS Perf
- Fruit 05/11/03 2881 16.5-7.5 (+14-5 =5 ) 100.0 +42
- Bright 0.3a 2837 13.0-11.0 (+8 -6 =10) 100.0 -114
- Chess Tiger 2007.1 2816 18.5-5.5 (+15-2 =7 ) 100.0 +27
- Frenzee Feb08 2763 20.5-3.5 (+18-1 =5 ) 100.0 +57
- SmarThink 1.00 32-bit 2763 18.0-6.0 (+14-2 =8 ) 100.0 -43
- Alaric 707 2760 19.0-5.0 (+16-2 =6 ) 100.0 -5
- Movei 00.8.438(10 10 10) 2737 19.5-4.5 (+18-3 =3 ) 100.0 +17
- E.T. Chess 13.01.08 2732 11.0-1.0 (+10-0 =2 ) 100.0 +114
- Delfi 5.2 2711 18.5-5.5 (+16-3 =5 ) 100.0 -64
- Frenzee Dec07 2704 11.0-1.0 (+10-0 =2 ) 100.0 +66
- Pro Deo 1.6(2007-10-11) 2703 20.5-3.5 (+19-2 =3 ) 100.0 +25
- Cerebro 3.01c 2577 8.0-0.0 (+8 -0 =0 ) 100.0 +INF