Who should compete in Michael's Tournament?

[overtond]

All sounds rather too sanctimonious to me - I wonder if John McCarthy would be that bothered that folk had taken his original Alpha-Beta idea and improved upon it or whether he would be very gratified to know that his idea in 1956 {the fundamentals anyway} is still used by many today?

[xsadar]

All sounds rather too sanctimonious to me - I wonder if John McCarthy would be that bothered that folk had taken his original Alpha-Beta idea and improved upon it or whether he would be very gratified to know that his idea in 1956 {the fundamentals anyway} is still used by many today?

It seems you've completely missed the point. Nothing sanctimonious about it. It's only about the fairness of a competition, particularly one where prize money is involved. Allowing clones in a computer vs. computer tournament is no more fair than allowing me to enter myself 100 times in a human vs. human tournament and varying my style based on which me I'm being at the time. Nor would it be fair if somebody else got the prize money rather than myself for me winning a chess tournament.

Now, of course, you're comment about alpha-beta is completely off topic. There's a huge difference between ideas and implementation.

Not sure why I'm commenting though, this discussion's going nowhere fast.

[marcelk]

All sounds rather too sanctimonious to me - I wonder if John McCarthy would be that bothered that folk had taken his original Alpha-Beta idea and improved upon it or whether he would be very gratified to know that his idea in 1956 {the fundamentals anyway} is still used by many today?

Now, of course, you're comment about alpha-beta is completely off topic. There's a huge difference between ideas and implementation.

Not sure why I'm commenting though, this discussion's going nowhere fast.

That is why I talk about using unlicensed work instead of cloning. The first term has a legal meaning, the second is just as meaningless as "pirating". Some indeed miss the point about what is the difference between taking an idea or taking a realization of an idea. Such discussion goes nowhere. I have seen otherwise respectable persons have difficulty grasping the problems of allowing unlicensed code to enter tournaments and being proud of their ignorance.

Ofttimes the bigger ELO contribution is in a realization (a debugged algorithm, a balanced set of parameters, etc) not so much in the idea itself. Ideas are worthless in a sense. You see a lot of them being posted here without being backed up by a realization to proof that it has any merit, and they all fade away.

But even without considering what provides more ELO, the existing IP laws of many countries (copyright, patents, DMCA, dislike them or not) do quite a good job outlining what can't be done.

Copyright infringement means copying a realization without permission or constructing a derived work from it. Reverse compiling a closed-source commercial program falls into this category (or under DMCA otherwise). Tuning a program to match as much as possible the output of closed-source commercial program is probably a DMCA violation as well.

Taking an idea is almost always permitted by IP laws, except when the idea is patented. Luckily, hardly any chess engine idea is patented because the costs are so high. If an idea is reverse engineered from observing and analyzing the behavior of a closed-source program, that is still fine: Such trade secrets are not protected by copyright or patent law. (DMCA might have something to say about this though).

Taking a legit GPL'ed program as a starting point to create another engine should probably be permitted, unless the tournament or author poses additional restrictions on such use. Although I can see and feel the frustration of engine programmers getting beaten by a Fruit-based program, I also think that any engine that is still weaker than those open source engines only has to blame himself for not outperforming them.