I don't see the analysis of Houdini in this Championship?

[Uri Blass]

I fail to see why the use of the engine would need special permission to be used in a website.

Because it is not be allowed by the license agreement:

Subject to the terms and conditions of this agreement, permission is granted
to install and execute the Houdini Chess Engine (the "Software") and use the
associated files for personal use. This license does not grant any right of
additional use other than the above.

This Software or the accompanying files may not be reproduced in any way
without prior written permission from the copyright holder.
You may not rent, lease, assign, transfer or re-distribute to the public the
Software. You may not modify the Software or merge all or any part of them in
another program, nor reverse engineer, decompile or disassemble the Software.

Publishing Houdini analysis on a commercial web site is not covered by "personal use".

Robert

I think that complaining about it is a mistake even if technically it is illegal.

You can only earn from publishing Houdini's analysis on a commercial web site because it is possible that people who never heard about houdini are going to hear about it and buy it.

If you do not like it then the commercial web site may use Komodo or Rybka to analyze the games and you are going to lose customers.

I also see no difference between posting houdini's analysis in a commercial web site and posting it in this forum and I see no justified reason to make one of them legal and one of them illegal.

[chrisw]

If testers in this forum and other forums test Houdini and post the results and allow games to be downloaded is that breaking the copyright if copyright holder did not approve it first.

Please stop trolling.

If one don't understand the difference between posting a game and publishing an Engine analysis on a website for the most important event in Chess, then

Not specifically referring to the above but there seems to be some confusion here. The "personal use" concept comes from the music and video industry, where broadcast to an audience is excluded by the "personal use" term.

But video and audio is different to software, the output of a video/audio CD is the copyright material, the song, the film or whatever and you would have no licence to distribute the copyright material to others again and again. The output of a program however is not copyright to the developer, his copyright extends to the code and support data only. Thus Robert Houdart, or anyone else, has no right to prevent use of the program output in any way, "personal use" terms or not.

This *might* (I say might because I'm not a specialist of compared law, so my knowledge of the anglo-saxon law is very superficial) be true in anglo-saxons countries, Chris, but not in other countries. For example in continental Europe the copyright laws for software are exactly the same as the ones used for books and movies/songs. And in Belgium, where I believe Robert lives, the law is nearly a duplicate of the French one (Belgium's law is based on Napoléon's civil code, just like France). They are divided in two categories:

- moral rights (those are actually attenuated, when compared to those for books/movies/songs, but we won't discuss them here).
- patrimonial rights. Those ones are subdivided into:
* reproduction rights
* representation rights
* traduction, adaptation rights
* selling rights

In our case here, the problem concerns the representation rights. They are often defined as "the communication of the work to the public by any mean" (ex: demonstrating a software during a fair, conferences, and so on...). This right is only suspended for the "private representations, and exclusively inside the family circle." This is not exactly what this FIDE event is.

Other accepted exceptions are:
- analyses and quotations of the work;
- press reviews;
- broadcast to the public for pure information about the work;
- parody, mockery.

So the question is: is broadcasting Houdini's output to the public a representation of Houdini? I don't think so. An illegal representation of Houdini would be to let the public of an event use Houdini on public computers.

ok, but another example ....

suppose we write some software, a simple algorithm, it takes any page of text, extracts the 1st, 13th, 25th, 19th etc etc according to a stored formula, 37th word and concatenates them into a sentence.

output = blaha blahb blahz la di da

My contention is that we have copyright over the program but not over the output. Simplistically we have:

program operates on [huge sample of all the world's texts] and outputs a string of data according to an algorithm.

Likewise, chess program operates on [huge sample of possible chess positions] and outputs a string of data according to an algorithm.

In neither case can the programmer predict the output data string, because he cannot personally deal with all the possible input situations. The programmer cannot lay copyright claim over a huge subset of possible output data strings that he doesn't know about and can't predict and where he also has no copyright over part of the data that goes into creating the output, namely the massively large possible input situations.

Yes, you are right. Kant proposed a simple philosophical tool to see if an idea is right or wrong: universalization. To check if an idea is absurd or not, push it to its limits. Here, for instance, we could generalize your program in order to produce a totally random (in length and content) string of characters. If this result could be copyrighted then we could claim a copyright over any future book or article (because we could say that it was produced using our software).

But here the questions was legally even simpler: the question is not "Does the FIDE has the right to publish Houdini's output to the public", but more exactly "does the FIDE has the right to produce a public output with Houdini?" And the answer to this second question is "legally, yes". Although it depends on Houdini's user license that the FIDE agreed to. Maybe Robert Houdard specifically forbids to publish Houdini's outputs in public places?

Sorry, but my French is not yet good enough to understand the difference between "Does the FIDE has the right to publish Houdini's output to the public" and "does the FIDE has the right to produce a public output with Houdini?"

Of course you would need permission to publish the output if the output contained proprietary material. For example a science fiction game might include moon map artwork and moon fighter character artwork, so publishing images containing these could be a copyright infringement. But I am assuming the Houdini "output" consists of alphanumerics only, e2e4 and 0.56 and so on. Robert Houdart cannot claim copyright over those.

[syzygy]

This *might* (I say might because I'm not a specialist of compared law, so my knowledge of the anglo-saxon law is very superficial) be true in anglo-saxons countries, Chris, but not in other countries. For example in continental Europe the copyright laws for software are exactly the same as the ones used for books and movies/songs.

But books and movies/songs do not produce output.

The analysis output by Houdini or any other engine is not copyrighted at all, so it certainly not covered by the copyright on Houdini. The output does not form a "work" in the sense of copyright law, because it lacks creativity.

Another thing is that Houdini's license agreement might forbid the licensee to use the engine for a particular purpose. However, in Europe a software license cannot forbid the lawful acquirer of a copy of the software to use the program in accordance with its intended purpose.

I would say that one of the intended purposes of a chess engine is to use it for analysing chess games. So producing the analysis is legal. What happens afterwards with the analysis is not for the engine's copyright holder to decide.

So the question is: is broadcasting Houdini's output to the public a representation of Houdini? I don't think so. An illegal representation of Houdini would be to let the public of an event use Houdini on public computers.

Aha, we in fact do agree

Btw, there is some reminiscense with C-393/09 in which the ECJ decided that the copyright on a computer program does not cover its graphical user interface or the display of the graphical user interface. A graphical user interface can however itself constitute a copyrighted "work". Such is however not the case for analysis output by a chess engine.

If a chess engine is used by a human as a tool in annotating a game, then that annotation could be copyrighted, but the author and rightholder would be this human, not the engine's author.

[JuLieN]

If testers in this forum and other forums test Houdini and post the results and allow games to be downloaded is that breaking the copyright if copyright holder did not approve it first.

Please stop trolling.

If one don't understand the difference between posting a game and publishing an Engine analysis on a website for the most important event in Chess, then

Not specifically referring to the above but there seems to be some confusion here. The "personal use" concept comes from the music and video industry, where broadcast to an audience is excluded by the "personal use" term.

But video and audio is different to software, the output of a video/audio CD is the copyright material, the song, the film or whatever and you would have no licence to distribute the copyright material to others again and again. The output of a program however is not copyright to the developer, his copyright extends to the code and support data only. Thus Robert Houdart, or anyone else, has no right to prevent use of the program output in any way, "personal use" terms or not.

This *might* (I say might because I'm not a specialist of compared law, so my knowledge of the anglo-saxon law is very superficial) be true in anglo-saxons countries, Chris, but not in other countries. For example in continental Europe the copyright laws for software are exactly the same as the ones used for books and movies/songs. And in Belgium, where I believe Robert lives, the law is nearly a duplicate of the French one (Belgium's law is based on Napoléon's civil code, just like France). They are divided in two categories:

- moral rights (those are actually attenuated, when compared to those for books/movies/songs, but we won't discuss them here).
- patrimonial rights. Those ones are subdivided into:
* reproduction rights
* representation rights
* traduction, adaptation rights
* selling rights

In our case here, the problem concerns the representation rights. They are often defined as "the communication of the work to the public by any mean" (ex: demonstrating a software during a fair, conferences, and so on...). This right is only suspended for the "private representations, and exclusively inside the family circle." This is not exactly what this FIDE event is.

Other accepted exceptions are:
- analyses and quotations of the work;
- press reviews;
- broadcast to the public for pure information about the work;
- parody, mockery.

So the question is: is broadcasting Houdini's output to the public a representation of Houdini? I don't think so. An illegal representation of Houdini would be to let the public of an event use Houdini on public computers.

ok, but another example ....

suppose we write some software, a simple algorithm, it takes any page of text, extracts the 1st, 13th, 25th, 19th etc etc according to a stored formula, 37th word and concatenates them into a sentence.

output = blaha blahb blahz la di da

My contention is that we have copyright over the program but not over the output. Simplistically we have:

program operates on [huge sample of all the world's texts] and outputs a string of data according to an algorithm.

Likewise, chess program operates on [huge sample of possible chess positions] and outputs a string of data according to an algorithm.

In neither case can the programmer predict the output data string, because he cannot personally deal with all the possible input situations. The programmer cannot lay copyright claim over a huge subset of possible output data strings that he doesn't know about and can't predict and where he also has no copyright over part of the data that goes into creating the output, namely the massively large possible input situations.

Yes, you are right. Kant proposed a simple philosophical tool to see if an idea is right or wrong: universalization. To check if an idea is absurd or not, push it to its limits. Here, for instance, we could generalize your program in order to produce a totally random (in length and content) string of characters. If this result could be copyrighted then we could claim a copyright over any future book or article (because we could say that it was produced using our software).

But here the questions was legally even simpler: the question is not "Does the FIDE has the right to publish Houdini's output to the public", but more exactly "does the FIDE has the right to produce a public output with Houdini?" And the answer to this second question is "legally, yes". Although it depends on Houdini's user license that the FIDE agreed to. Maybe Robert Houdard specifically forbids to publish Houdini's outputs in public places?

Sorry, but my French is not yet good enough to understand the difference between "Does the FIDE has the right to publish Houdini's output to the public" and "does the FIDE has the right to produce a public output with Houdini?"

Of course you would need permission to publish the output if the output contained proprietary material. For example a science fiction game might include moon map artwork and moon fighter character artwork, so publishing images containing these could be a copyright infringement. But I am assuming the Houdini "output" consists of alphanumerics only, e2e4 and 0.56 and so on. Robert Houdart cannot claim copyright over those.

Excuse my french The second question precedes the first one. I should have simplified it that way: "Does the FIDE has the right to use Houdini to produce an output intended for the public?" So it's a more basic question than "Does the FIDE has the right to publish Houdini's output to the public?", because we're asking ourselves this question before even using Houdini, whereas the second question comes after Houdini has produced an output. If the question to the basic question is "yes it is legal, and yes the license doesn't prevent it", then you don't even need to ask yourself the other question. And yes, it is legal. Now for the license I don't know: I don't own Houdini so I've not taken a look to its license.

We agree on this question, Chris, so you don't need to bring scifi in the discussion. YOU are the scifi.

EDIT: Yes Ronald, we agree.

[gleperlier]

You know what guys ? We succeeded in letting this thread calm and polite !

Congratulations to all

[Graham Banks]

...Excuse my french......

Haha.

[Don]

If testers in this forum and other forums test Houdini and post the results and allow games to be downloaded is that breaking the copyright if copyright holder did not approve it first.

Please stop trolling.

If one don't understand the difference between posting a game and publishing an Engine analysis on a website for the most important event in Chess, then

Not specifically referring to the above but there seems to be some confusion here. The "personal use" concept comes from the music and video industry, where broadcast to an audience is excluded by the "personal use" term.

But video and audio is different to software, the output of a video/audio CD is the copyright material, the song, the film or whatever and you would have no licence to distribute the copyright material to others again and again. The output of a program however is not copyright to the developer, his copyright extends to the code and support data only. Thus Robert Houdart, or anyone else, has no right to prevent use of the program output in any way, "personal use" terms or not.

This *might* (I say might because I'm not a specialist of compared law, so my knowledge of the anglo-saxon law is very superficial) be true in anglo-saxons countries, Chris, but not in other countries. For example in continental Europe the copyright laws for software are exactly the same as the ones used for books and movies/songs. And in Belgium, where I believe Robert lives, the law is nearly a duplicate of the French one (Belgium's law is based on Napoléon's civil code, just like France). They are divided in two categories:

- moral rights (those are actually attenuated, when compared to those for books/movies/songs, but we won't discuss them here).
- patrimonial rights. Those ones are subdivided into:
* reproduction rights
* representation rights
* traduction, adaptation rights
* selling rights

In our case here, the problem concerns the representation rights. They are often defined as "the communication of the work to the public by any mean" (ex: demonstrating a software during a fair, conferences, and so on...). This right is only suspended for the "private representations, and exclusively inside the family circle." This is not exactly what this FIDE event is.

Other accepted exceptions are:
- analyses and quotations of the work;
- press reviews;
- broadcast to the public for pure information about the work;
- parody, mockery.

So the question is: is broadcasting Houdini's output to the public a representation of Houdini? I don't think so. An illegal representation of Houdini would be to let the public of an event use Houdini on public computers.

ok, but another example ....

suppose we write some software, a simple algorithm, it takes any page of text, extracts the 1st, 13th, 25th, 19th etc etc according to a stored formula, 37th word and concatenates them into a sentence.

output = blaha blahb blahz la di da

My contention is that we have copyright over the program but not over the output. Simplistically we have:

program operates on [huge sample of all the world's texts] and outputs a string of data according to an algorithm.

Likewise, chess program operates on [huge sample of possible chess positions] and outputs a string of data according to an algorithm.

In neither case can the programmer predict the output data string, because he cannot personally deal with all the possible input situations. The programmer cannot lay copyright claim over a huge subset of possible output data strings that he doesn't know about and can't predict and where he also has no copyright over part of the data that goes into creating the output, namely the massively large possible input situations.

Yes, you are right. Kant proposed a simple philosophical tool to see if an idea is right or wrong: universalization. To check if an idea is absurd or not, push it to its limits. Here, for instance, we could generalize your program in order to produce a totally random (in length and content) string of characters. If this result could be copyrighted then we could claim a copyright over any future book or article (because we could say that it was produced using our software).

But here the questions was legally even simpler: the question is not "Does the FIDE has the right to publish Houdini's output to the public", but more exactly "does the FIDE has the right to produce a public output with Houdini?" And the answer to this second question is "legally, yes". Although it depends on Houdini's user license that the FIDE agreed to. Maybe Robert Houdard specifically forbids to publish Houdini's outputs in public places?

For example if I test a position using Houdini and then I post the output, it's no longer personal use because I am making the analysis public.

I don't agree with the philosophical tool of pushing it to the limit. I'm not refuting your argument but what my understand of what Kant said from your description. Is eating good for you? Of course it is, you will die if you don't eat. Push it to the limit and you will die of gluttony in one day. Some medications can save your life - but overdoes on them and they will kill you.

[JuLieN]

I don't agree with the philosophical tool of pushing it to the limit. I'm not refuting your argument but what my understand of what Kant said from your description. Is eating good for you? Of course it is, you will die if you don't eat. Push it to the limit and you will die of gluttony in one day. Some medications can save your life - but overdoes on them and they will kill you.

Kant's tool was used in Critique of Pure Reason and only meant to be used in the field of morality. Now you can extend the use if this tool to fields that are close to morality (politics, law, human sciences, and so on...) but it probably won't be a good idea with hard sciences. As for myself I only use it when I need to assess my political or moral opinions or other people's.

For example if I test a position using Houdini and then I post the output, it's no longer personal use because I am making the analysis public.

Yes, it's still personal use: you're the only one using the software (regarding the software's output: see above discussion). Now if you have a penniless club player of your friends who asks you to borrow your computer and Houdini to prepare for his next tournament, this is not a personal use of the software anymore. Although there is a tolerance, as yet stated: the "family circle". Usually that will mean at your place and with your relatives.

EDIT: BTW, In forgot something VERY important that simplifies the debate even more: only the "oeuvres de l'esprit" (the mind's works) are protected by the law, and every time a court (at least in France) had to decide if a work automatically produced by a computer could be protected by copyright law they decided against it. I remember for instance, when I was in law school, an example that had amused me: a software that could generate new tunes automatically. Those tunes can't be protected, because they're not produced by a human mind. Obviously it's the same thing for a chess engine's output.

This might change the day we create an artificial consciousness: this entity could claim a legal identity and thus get its work protected.

[chrisw]

If testers in this forum and other forums test Houdini and post the results and allow games to be downloaded is that breaking the copyright if copyright holder did not approve it first.

Please stop trolling.

If one don't understand the difference between posting a game and publishing an Engine analysis on a website for the most important event in Chess, then

Not specifically referring to the above but there seems to be some confusion here. The "personal use" concept comes from the music and video industry, where broadcast to an audience is excluded by the "personal use" term.

But video and audio is different to software, the output of a video/audio CD is the copyright material, the song, the film or whatever and you would have no licence to distribute the copyright material to others again and again. The output of a program however is not copyright to the developer, his copyright extends to the code and support data only. Thus Robert Houdart, or anyone else, has no right to prevent use of the program output in any way, "personal use" terms or not.

This *might* (I say might because I'm not a specialist of compared law, so my knowledge of the anglo-saxon law is very superficial) be true in anglo-saxons countries, Chris, but not in other countries. For example in continental Europe the copyright laws for software are exactly the same as the ones used for books and movies/songs. And in Belgium, where I believe Robert lives, the law is nearly a duplicate of the French one (Belgium's law is based on Napoléon's civil code, just like France). They are divided in two categories:

- moral rights (those are actually attenuated, when compared to those for books/movies/songs, but we won't discuss them here).
- patrimonial rights. Those ones are subdivided into:
* reproduction rights
* representation rights
* traduction, adaptation rights
* selling rights

In our case here, the problem concerns the representation rights. They are often defined as "the communication of the work to the public by any mean" (ex: demonstrating a software during a fair, conferences, and so on...). This right is only suspended for the "private representations, and exclusively inside the family circle." This is not exactly what this FIDE event is.

Other accepted exceptions are:
- analyses and quotations of the work;
- press reviews;
- broadcast to the public for pure information about the work;
- parody, mockery.

So the question is: is broadcasting Houdini's output to the public a representation of Houdini? I don't think so. An illegal representation of Houdini would be to let the public of an event use Houdini on public computers.

ok, but another example ....

suppose we write some software, a simple algorithm, it takes any page of text, extracts the 1st, 13th, 25th, 19th etc etc according to a stored formula, 37th word and concatenates them into a sentence.

output = blaha blahb blahz la di da

My contention is that we have copyright over the program but not over the output. Simplistically we have:

program operates on [huge sample of all the world's texts] and outputs a string of data according to an algorithm.

Likewise, chess program operates on [huge sample of possible chess positions] and outputs a string of data according to an algorithm.

In neither case can the programmer predict the output data string, because he cannot personally deal with all the possible input situations. The programmer cannot lay copyright claim over a huge subset of possible output data strings that he doesn't know about and can't predict and where he also has no copyright over part of the data that goes into creating the output, namely the massively large possible input situations.

Yes, you are right. Kant proposed a simple philosophical tool to see if an idea is right or wrong: universalization. To check if an idea is absurd or not, push it to its limits. Here, for instance, we could generalize your program in order to produce a totally random (in length and content) string of characters. If this result could be copyrighted then we could claim a copyright over any future book or article (because we could say that it was produced using our software).

But here the questions was legally even simpler: the question is not "Does the FIDE has the right to publish Houdini's output to the public", but more exactly "does the FIDE has the right to produce a public output with Houdini?" And the answer to this second question is "legally, yes". Although it depends on Houdini's user license that the FIDE agreed to. Maybe Robert Houdard specifically forbids to publish Houdini's outputs in public places?

For example if I test a position using Houdini and then I post the output, it's no longer personal use because I am making the analysis public.

I don't agree with the philosophical tool of pushing it to the limit. I'm not refuting your argument but what my understand of what Kant said from your description. Is eating good for you? Of course it is, you will die if you don't eat. Push it to the limit and you will die of gluttony in one day. Some medications can save your life - but overdoes on them and they will kill you.

Eating and taking medication are not philosophical ideas and their use in this topic is therefore as apples to cheese.

Reverting to Houdini, it appears to have been conclusively demonstrated on open chess in the past days that it is perfectly legal to create a functionally equivalent version of a program (assuming non usage of same source code). So, we can create a functionally equivalent version of Houdini, legally. We can of course then do exactly whatever we want with the outputs of this functionally equivalent program we just created (so long as the outputs do not contain propriatory graphics, for example, just alphanumerics).

From this we can argue that neither Houdini, not any functional equivalent of Houdini, is going to be able to claim any form of copyright on its output, since anybody can create this output, legally.

[JuLieN]

Eating and taking medication are not philosophical ideas and their use in this topic is therefore as apples to cheese.

Reverting to Houdini, it appears to have been conclusively demonstrated on open chess in the past days that it is perfectly legal to create a functionally equivalent version of a program (assuming non usage of same source code). So, we can create a functionally equivalent version of Houdini, legally. We can of course then do exactly whatever we want with the outputs of this functionally equivalent program we just created (so long as the outputs do not contain propriatory graphics, for example, just alphanumerics).

From this we can argue that neither Houdini, not any functional equivalent of Houdini, is going to be able to claim any form of copyright on its output, since anybody can create this output, legally.

Yep. Sorry you had to write that before I added my precision to my previous post:

EDIT: BTW, In forgot something VERY important that simplifies the debate even more: only the "oeuvres de l'esprit" (the mind's works) are protected by the law, and every time a court (at least in France) had to decide if a work automatically produced by a computer could be protected by copyright law they decided against it. I remember for instance, when I was in law school, an example that had amused me: a software that could generate new tunes automatically. Those tunes can't be protected, because they're not produced by a human mind. Obviously it's the same thing for a chess engine's output.

This might change the day we create an artificial consciousness: this entity could claim a legal identity and thus get its work protected.

Regarding the other point ("it is perfectly legal to create a functionally equivalent version of a program (assuming non usage of same source code)."), true also. Although remember that reverse-engineering is forbidden to create a competing product (but your "assuming non usage of same source code" yet covers this objection. ).