In Summary

[garybelton]

At a minimum, for example:

"HIARCS by Mark John Uniacke (based om multiple ideas from the public domain)". Even better, not to have an "Author" tag at all, but instead to have a coder tag:

uci
Coder Mark John Uniacke

In the case of a chess engine, the word "coder" is much more appropriate than "author", given the use of the multiple ideas in the program that did not originate from the brain of the chess engine coder(s). With these subtle changes, the case of plagiarism (as defined by Wikipedia) cannot be made.

The most accurate tag of those posted is actually Fires!

[hgm]

I think that you are overlooking that WCCC rules have to be more strict than copyright rule.

Copying ideas is OK according to copyright law
Copying ideas might not be OK for WCCC

At least, if the gray area of eval parameters and engine tuning are considered ideas. (And, according to copyright law, it would be allowed to copy them, so heaping them with the ideas makes sense.)

We don't want to have 5 independently coded but identically playing engines participating in the WCCC. (Which have the same move ordering, the same extensions, the same reductions, the same eval terms, the same eval parameters... But totally different code and data structures.) The WCCC is not intended to be an instrument for measuring who could most efficiently implement the same algorithm, or which of its implementers is simply most lucky. There could be contests for that too, and they might even be interesting and beneficial to the computer Chess community. But it is ICGA's choice for the WCCC to not be such an instrument, and it is their privilege to decide that, because it is their tournament. If others want to organize an IppoCloners or FruitMash Contest, they are free to do so. And they could disqualify anyone for making an entry that did not look enough like Ippo, because they cheated by using an _original_ evaluation term that improved Elo...

So we somehow have to better define the gray area (not gray in copyright law. but according to competition rules) of implementing 'ideas'. Some ideas are public-domain knowledge, or even mathematically proven optima (minimax, alpha-beta), and it would be silly to force anyone to stray from a generally known optimum just to count as original. This is one extreme of the spectrum. The other extreme is ripping the complete evaluation from a specific engine, and (after recoding) use it in your own. I would say this should not be allowed in competition rules even if the engine in question was open source, like Fruit. Completely recoded Fruits would still not be welcomed.

[garybelton]

I don't disagree with anything you say there HG and I wish the ICGA good luck in better defining the grey areas and their future tournaments. My post, which you responded to, was pointing out something completely different. By the Wikipedia definition all chess engine authors are plagiarists. This is important for any defense should this ever go to court.

[hgm]

Well, this can never go to court, as it is an internal ICGA matter, and even if it would not, I would expect neither the court nor ICGA to recognize Wikipedia's authority on anything...

[garybelton]

I agree the ICGA banning decision won't go to court, I am talking about a defamation/libel case. The Wikipedia entry is built from a number of dictionary sources so it could hold up.

If the defamation suit gets funded and the evidence comes out in court the defense would have a field day. Jonnie Cochrane has died now, but it would go something like:

1. All chess engine authors are plagiarists
2. The best chess engine author had his program reverse engineered so he withdrew making improvements in public, but carried on making improvements in private
3. Chess engine authors wanted to see what these private improvements were, so that they could plagiarise them, but couldn't get their hands on the executable to reverse engineer, or source code
4. So, they demanded that the Rybka author provide the Rybka source code under an ICGA ruling for originality
5. The author failed to comply and so was banned, stripped of four world titles and publicly humiliated
6. Some of his titles went directly to the people that banned him (a further motive and conflict of interest)
7. At no stage could the prosecution put up two pieces of code side by side, one from Fruit, one from Rybka and say "look they are the same". As a result of this inability, the defense lawyers would claim that the evidence against the Rybka author for the ICGA events is only circumstantial

The defense would also throw in accusations of a personal vendetta being carried out by a senior computer scientist in his twilight years who considers that he didn't get enough credit and has been seriously eclipsed by a younger programmer, especially in the area of parallel programming, in which the senior computer scientist made his name. The defense would show that this strong character railroaded many into attacking the defendant and would make reference to forum links to present him as an unbalanced, aggressive and immature (playing online first person shooter games at aged 60+) individual. As you yourself mentioned, these lawyer guys go are mean for the jugular.

This defense would have a good chance of winning, it doesn't matter about what/what not really happended. All the Rybka author needs to do is find a glove a couple of sizes too small to wear in court.

[hgm]

I think you are mixing up "defense" and "plaintiff" now. Surely you don't expect Rybka to be accused of libel/defamation?

[garybelton]

Well kind of. It's been said that as part of a defamation case, the evidence (and even more evidence) would come out against the Rybka author so they would need to have a defense to that before continuing as an accuser for libel damages.

If an anonymous Sheik is unable to foot the legal bill and his own pockets are not deep enough then I recommend to the Rybka author to contact some "no-win, no-fee" legal organizations in Poland. This has gone far outside of the relatively small computer chess community, the damages from a defamation/libel suit in this case would be an easy pay day for them, unfortunately knowing how they work they would probably end up with the lions share of the recovered damages themselves.

On another note, Winboard is great by the way, the best chess gui out there, thank you for keeping up work on it after Tim Mann & co.

[hgm]

If they started the lawsuit, they would still be the plaintiff. No matter if there would be a counter-claim.

If you are rich enough I guess you could sue, and eventually break anyone, no matter how much in the wrong you are. There is not much incentive to do that in this case, though, because there is not really anything to gain. In the hypothetical and unlikely situation that ICGA would be found guilty of defamation, having to pay damages, they would simply immediately apply for bankrupcy, and no cent would ever be payed. They can hardly scape enough money together to afford printing of their Journal every 3 months... The trophy is wood, not much from melting that either... The people involved would start a new organization, ICGB, and that would be the end of it.

So yes, a fabulously wealthy oil sheik could probably buy a victory in court. A quite hollow victory. Why would anyone want to poor his money down the drain that way? The whole idea seems an utterly ridiculous day dream...

[garybelton]

they would simply immediately apply for bankrupcy, and no cent would ever be payed.

It may be possible to pursue the ICGA board for their personal assets, I don't know how the ICGA is incorporated. In a successful outcome for the Rybka author, at the minimum a retraction of the ICGA statement and the news agencies who picked up the story could be forced.

The whole idea seems an utterly ridiculous day dream...

Depends if a Jury was present or not. Have you ever been to a court hearing involving a Jury and Expert Witnesses? This is how it would go, in this case:

1) Jury sees the experts from both sides present testimonies that disagree with eachother
2) Jury tries to assess the expert testimony
3) Jury has no clue what these experts (two compiler guys and two contract (GPL) guys in this case) are talking about
4) Jury gives them a point each as the compiler guys both have long beards (so must know what they are talking about) and the contract guys have good suits
5) Jury says "show us the copied code"
6) Jury notes that the exhibit shown looks nothing like the Fruit code they were shown previously
7) The Rybka author defense arguments are made (see previous post)
8) Jury wonders why a bunch of plagiarisers are accusing the best of them as a plagiariser when they are all clearly plagiarisers
9) Case is awarded to plaintiff and the Jury looks for lunch ...

If the Rybka author is going to sue I recommend a Jury hearing for his optimal outcome.

[hgm]

It may be possible to pursue the ICGA board for their personal assets,

That sounds like another desparate day dream. Suing board members is in general only possible in cases of gross mismanagement, violating their own, legally approved company rules, or the law.

I don't know how the ICGA is incorporated. In a successful outcome for the Rybka author, at the minimum a retraction of the ICGA statement and the news agencies who picked up the story could be forced.

That seems not be possible without attacking their decision to disquaify Rybka in the first place (which I thought you did not want to do, going for the defamation approach in stead). What exactly would they have to retract if the decision still stands? They would be ruled to make the same statement in a slightly more friendly wording?

Depends if a Jury was present or not. Have you ever been to a court hearing involving a Jury and Expert Witnesses? This is how it would go, in this case:

1) Jury sees the experts from both sides present testimonies that disagree with eachother
2) Jury tries to assess the expert testimony
3) Jury has no clue what these experts (two compiler guys and two contract (GPL) guys in this case) are talking about
4) Jury gives them a point each as the compiler guys both have long beards (so must know what they are talking about) and the contract guys have good suits
5) Jury says "show us the copied code"
6) Jury notes that the exhibit shown looks nothing like the Fruit code they were shown previously
7) The Rybka author defense arguments are made (see previous post)
8) Jury wonders why a bunch of plagiarisers are accusing the best of them as a plagiariser when they are all clearly plagiarisers
9) Case is awarded to plaintiff and the Jury looks for lunch ...

If the Rybka author is going to sue I recommend a Jury hearing for his optimal outcome.

This seems to have little to do with footing the bill for the lawyers. We don't have jury trials over here, so perhaps my understanding is incomplete, but I thought it was not the jury that has to pay for the trial... The day dream is that some Santa Clause would step in to buy a hollow victory at huge cost, out of pity for Vas.

Do you know where the ICGA is registered, and which jurisdiction would apply? Is a jury trial even possible in that jurisdiction?