creating an opening book

[Ferdy]

Building a book based on free strong engine is not cheating because everybody can take the engine and use it - playing field is even

How about non-free hardware?

Creating an opening book using an engine that is not yours, say a strong engine like SF to check every positions in the book is not cheating. Are we going to discuss the kind of HW used in creating such opening book and/or used by the engine in the analysis of the positions in the book?

To clarify, the question is what differentiates free programs from non-free programs in your judgement on this fairness matter, and how does that translate to hardware if it does. Given that Komodo (for example) is just as available for everyone as SF. And given that hardware is never free, but still a necessity to use even SF to make a book. It is not obvious how to reconcile that apparent contraction, hence the question.

The issue really is the use of engine that is not yours in building the book. The use of a free strong program say SF should not be considered cheating because everyone has access to it.
In competition the skill of the programmer should be tested as much as possible. If the author is able to implement SMP in his program that is fine, it is his skill, but using a single threaded engine with a high end cpu because he can afford it, is not fair when his opponents are also using a single threaded program but uses only a low end cpu.

And if someone has bought and paid for a chess program they should not be able to use it to analyse data for their opening book?

This is unfair if that program someone has purchased is stronger than the strongest free program available. The author is no longer competing at even level. You can still use that engine but the book created with the help of that engine is questionable when entered in the tournament or competition unless of course the other participants are also using that purchased engine then we are back in even level.

How is that different from purchasing the assistance of a grandmaster? Is that also unfair?

It is fine if the GM is using SF 6 for example or it uses its knowledge and experience alone in creating the opening book.

So GMs who use programs that they have purchased are doing something unfair?

For programmers testing their skills/abilities thru the programs that they create by participating in tournaments, using SF6 (that is free) to help build the opening book for their program is fine and rewarding.

[mvk]

In competition the skill of the programmer should be tested as much as possible. If the author is able to implement SMP in his program that is fine, it is his skill, but using a single threaded engine with a high end cpu because he can afford it, is not fair when his opponents are also using a single threaded program but uses only a low end cpu.

I disagree here then: Computer chess is a very cheap hobby. Even high-end CPUs (single core performance) are affordable, and can even be rented for as little as a few pennies for the duration of a game. Unaffordable hardware (clusters) still doen't give any real advantage if the program is the weaker part. See for example the wins of Houdini against the Rybka cluster. Money plays only a minor role. Why would someone be handicapped for using contemporary consumer grade hardware? That sounds unfair to me.

[mhull]

How is that different from purchasing the assistance of a grandmaster? Is that also unfair?

It is fine if the GM is using SF 6 for example or it uses its knowledge and experience alone in creating the opening book.

This is not logical. The GM is better than any program (commercial or free) at opening theory. The GM is a "commercial" asset. This is not different than purchasing a commercial program for the same purpose.

[Ferdy]

How is that different from purchasing the assistance of a grandmaster? Is that also unfair?

It is fine if the GM is using SF 6 for example or it uses its knowledge and experience alone in creating the opening book.

This is not logical. The GM is better than any program (commercial or free) at opening theory.

That was before, but now there are equally non GM that are also good in book preparations this is because of the engine help and some software/databases that are available - Aquarium, chessbase and others that are not commercial.

[Ferdy]

In competition the skill of the programmer should be tested as much as possible. If the author is able to implement SMP in his program that is fine, it is his skill, but using a single threaded engine with a high end cpu because he can afford it, is not fair when his opponents are also using a single threaded program but uses only a low end cpu.

I disagree here then: Computer chess is a very cheap hobby. Even high-end CPUs (single core performance) are affordable, and can even be rented for as little as a few pennies for the duration of a game. Unaffordable hardware (clusters) still doen't give any real advantage if the program is the weaker part. See for example the wins of Houdini against the Rybka cluster. Money plays only a minor role. Why would someone be handicapped for using contemporary consumer grade hardware? That sounds unfair to me.

So you disagree on this.

In competition the skill of the programmer should be tested as much as possible.

The competition here refers to computer chess.

[jdart]

Good hardware is not cheap. A modern top-line multi-CPU machine is somewhere between $5K and $20K. And in my experience there is no such thing as too much CPU power. If you have a lot of it, you can test things faster, you can play more online and offline test games, you can be more competitive in tournaments if that is your goal, etc. I have 3 big server/workstation class machines and a few consumer-grade quads. Most are running nearly 24x7. Sometimes I rent a big box in the cloud, too.

--Jon

[Joost Buijs]

Good hardware is not cheap. A modern top-line multi-CPU machine is somewhere between $5K and $20K. And in my experience there is no such thing as too much CPU power. If you have a lot of it, you can test things faster, you can play more online and offline test games, you can be more competitive in tournaments if that is your goal, etc. I have 3 big server/workstation class machines and a few consumer-grade quads. Most are running nearly 24x7. Sometimes I rent a big box in the cloud, too.

Right, in the past I spent so much money on hardware to be competitive I could have easily bought 2 big Mercedes cars for it.
I stopped buying multi-processor systems because the added performance is IMHO not worth the money you burn on it.
At the moment I have two 6 core and one 8 core consumer grade Intel machines which are more than enough for the things I want to do, ironically they are doing nothing most of the time.
Lately I used one of the old machines to calculate 6 men syzygy which took about a week, but I still have to find inspiration to implement it in my engine.

[mvk]

Good hardware is not cheap.

Correct, but that is not the argument. Good-enough hardware is cheap. A multicore board with 14nm xeons is overkill for 24/7 testing (absolute time doesn't matter there, only throughput), and the added value in tournaments for an amateur program is very little. And when in need, it could be rented, borrowed, sponsored. No need to own such things, until you are maybe within 50 elo of the top.

[Dann Corbit]

Good hardware is not cheap.

Correct, but that is not the argument. Good-enough hardware is cheap. A multicore board with 14nm xeons is overkill for 24/7 testing (absolute time doesn't matter there, only throughput), and the added value in tournaments for an amateur program is very little. And when in need, it could be rented, borrowed, sponsored. No need to own such things, until you are maybe within 50 elo of the top.

Right, but the argument seems to be that only free tools should be used in the name of fairness.

Purchased or rented computer systems might also fall under this cloud of unfair advantage.
Adequate PCs can be obtained for a few hundred dollars:
http://www.tigerdirect.com/applications ... =WEBBP2547

But that is still several times more expensive than a chess program like Houdini or Komodo.

I would argue that obtaining a truly equal playing field (utterly equal in all aspects) would be impossible. That does not mean that trying to get some measure of equality is a bad idea. Only that it is a mistake to assume that real equality can be reached.

I seem to recall that there have been uniform platform events in the past (which does obtain at least one measure of equality). Of course, this severely punishes (or even disqualifies) those programs designed for other platforms.

[mvk]

I have once proposed uniform platform competitions not to be limited by some particular hardware design (which I dislike very much for the reasons you give), but by cost of the system. Or by mass, something like that, there are many ways.