"Fissure opens in chess AI scene" (Rob Beschizza/B

[JuLieN]

Here is a (long) and interesting article about the Rybka case:

http://boingboing.net/2012/01/19/fissur ... -scen.html

Please, before commenting, be sure to:

1- read the article up to its very end, and
2- don't go personal, do not insult, don't be aggressive or ironical, stay calm and be tolerant to other people's opinion.

Thank you.

[Don]

Here is a (long) and interesting article about the Rybka case:

http://boingboing.net/2012/01/19/fissur ... -scen.html

Please, before commenting, be sure to:

1- read the article up to its very end, and
2- don't go personal, do not insult, don't be aggressive or ironical, stay calm and be tolerant to other people's opinion.

Thank you.

Interesting article and I think the author tried to present both side fairly but did not entirely succeed. The problem is that it tended to focus on the algorithms way too much, making it seem as if the debate was about that. For example even in the second paragraph this comes up:

Debate centers on chess-playing algorithms found both in certain versions of Rybka and another program, Fruit.

However they did get a lot of this right.

[JuLieN]

Here is a (long) and interesting article about the Rybka case:

http://boingboing.net/2012/01/19/fissur ... -scen.html

Please, before commenting, be sure to:

1- read the article up to its very end, and
2- don't go personal, do not insult, don't be aggressive or ironical, stay calm and be tolerant to other people's opinion.

Thank you.

Interesting article and I think the author tried to present both side fairly but did not entirely succeed. The problem is that it tended to focus on the algorithms way too much, making it seem as if the debate was about that. For example even in the second paragraph this comes up:

Debate centers on chess-playing algorithms found both in certain versions of Rybka and another program, Fruit.

However they did get a lot of this right.

I had the exact same feeling about this article: impressive synthesis work (who is Rob Beschizza, btw? He is very well informed!), and a true attempt at being fair. But with an enormous mistake, as the ICGA doesn't accuse Vas of borrowing ideas from Fruit, but instead of having copied large portions of it verbatim, and hence having violated copyright laws.

Still, providing people are aware of this mistake, this article is a good introduction to the case.

[Sylwy]

...... do not insult, don't be aggressive or ironical, stay calm and be tolerant to other people's opinion.

Thank you.

OK - Maestro Julien !

Prédateur number 2.2.1 , please ?

Thank you !

Have a nice weekend !

SilvianR

[hgm]

I think this whole controversy is mostly fueled by lack of understanding by laymen about what the various level of abstraction in computer programming are, and what is exactly allowed in which context. So there is this ever recurring mixup of the various terms:

ideas
algorithms
implementations
souce code

Both Riis and Vas seem to be completely off-the-mark here too: the ICGA verdict is based on likeness at the implementation level that beyond reasonable doubt can only be obtained by copying (which is illegal under ICGA rules). Vas defense is that there was no proof of source-code copying (and claims he did not do that), while Riis defense is that everyone uses the same algorithms. Both are totally irrellevant. It is like defending killing a pedestrian on a zebra crossing in a road accident by either "but my car was properly checked by a mechanic just a week before" or "but many people drive on that same road".

Furthermore, it is often not recognized that a 'copying' is a process, rather than a state. (And Riis arguments also seem to suffer from that.) If I ask 100 people to write down how much 6 times 7 is, some of them would write '42'. Yet one 42 would in most cases not be a copy of any other of the 42s. That solely depends on how the person writing it down acted. Did he consult his memory for the table of multiplication, or did he look on his neighbor's paper to see what he wrote down, and then simply wrote down the same?

[marcelk]

I had the exact same feeling about this article: impressive synthesis work (who is Rob Beschizza, btw? He is very well informed!), and a true attempt at being fair. But with an enormous mistake, as the ICGA doesn't accuse Vas of borrowing ideas from Fruit, but instead of having copied large portions of it verbatim, and hence having violated copyright laws.

This shows the matter is difficult: ICGA has/had no say about copyright (only courts have a say about that), but about their own rule #2 which is somewhat different and in some aspects a lot more restrictive than copyright.

[Don]

I think this whole controversy is mostly fueled by lack of understanding by laymen about what the various level of abstraction in computer programming are, and what is exactly allowed in which context. So there is this ever recurring mixup of the various terms:

ideas
algorithms
implementations
souce code

Both Riis and Vas seem to be completely off-the-mark here too: the ICGA verdict is based on likeness at the implementation level that beyond reasonable doubt can only be obtained by copying (which is illegal under ICGA rules). Vas defense is that there was no proof of source-code copying (and claims he did not do that), while Riis defense is that everyone uses the same algorithms. Both are totally irrellevant. It is like defending killing a pedestrian on a zebra crossing in a road accident by either "but my car was properly checked by a mechanic just a week before" or "but many people drive on that same road".

Furthermore, it is often not recognized that a 'copying' is a process, rather than a state. (And Riis arguments also seem to suffer from that.) If I ask 100 people to write down how much 6 times 7 is, some of them would write '42'. Yet one 42 would in most cases not be a copy of any other of the 42s. That solely depends on how the person writing it down acted. Did he consult his memory for the table of multiplication, or did he look on his neighbor's paper to see what he wrote down, and then simply wrote down the same?

Very well put. I think you are right about most of what fuels the misunderstandings here.

There is yet another "disconnect" too that involves how people in general feel about the seriousness of the issue in the general case. For example 60 years ago many white people didn't see what the big deal was about discrimination that they thought didn't affect them personally - not really seeing the bigger picture that everyone is affected negatively.

[bob]

I had the exact same feeling about this article: impressive synthesis work (who is Rob Beschizza, btw? He is very well informed!), and a true attempt at being fair. But with an enormous mistake, as the ICGA doesn't accuse Vas of borrowing ideas from Fruit, but instead of having copied large portions of it verbatim, and hence having violated copyright laws.

This shows the matter is difficult: ICGA has/had no say about copyright (only courts have a say about that), but about their own rule #2 which is somewhat different and in some aspects a lot more restrictive than copyright.

"in what ways"???

[marcelk]

I had the exact same feeling about this article: impressive synthesis work (who is Rob Beschizza, btw? He is very well informed!), and a true attempt at being fair. But with an enormous mistake, as the ICGA doesn't accuse Vas of borrowing ideas from Fruit, but instead of having copied large portions of it verbatim, and hence having violated copyright laws.

This shows the matter is difficult: ICGA has/had no say about copyright (only courts have a say about that), but about their own rule #2 which is somewhat different and in some aspects a lot more restrictive than copyright.

"in what ways"???

In the ways explained by the TD during the ICGA triennial meeting and also in the ICGA journal.

For example in the way that entering a Fruit-based engine in an ICGA tournament is not a copyright infringement in any way, but it is a infringement of rule #2.

Or in the way that copyright allows fair-use, and rule #2 doesn't. (PST structures come to mind)

[bob]

I had the exact same feeling about this article: impressive synthesis work (who is Rob Beschizza, btw? He is very well informed!), and a true attempt at being fair. But with an enormous mistake, as the ICGA doesn't accuse Vas of borrowing ideas from Fruit, but instead of having copied large portions of it verbatim, and hence having violated copyright laws.

This shows the matter is difficult: ICGA has/had no say about copyright (only courts have a say about that), but about their own rule #2 which is somewhat different and in some aspects a lot more restrictive than copyright.

"in what ways"???

In the ways explained by the TD during the ICGA triennial meeting and also in the ICGA journal.

For example in the way that entering a Fruit-based engine in an ICGA tournament is not a copyright infringement in any way, but it is a infringement of rule #2.

Or in the way that copyright allows fair-use, and rule #2 doesn't. (PST structures come to mind)

Depends. IF the copier has permission, then it is not a copyright infringement. IF the copier has permission then he can enter in the ICGA. Example: Toga (cluster toga) in the past... So a copy CAN be entered, with permission of the original author, if the original author does not also enter.

Copyright doesn't allow "fair use" for you to copy the code and use it in ANOTHER program whether that program is sold or not.

So I am not sure how rule 2 is "more restrictive." It would seem to be about the same, as it is all about the original author and his permission to use his work.