Fruit vs. Toga poll

[Rolf]

Here are my other questions translated into easy mode without sarcasm:

1) Is it for you a difference if you would end on place 12 in CCT or someone else as operator/pretended owner of basically your engine?

2) Do you see a similarity in the status of FRUIT and then GLAURUNG?

3) Do you call for as many as possible Glaurung entries in CCT?

4) Do you like to see Toga entry and other Fruit clones?

5) (Please excuse the clarity in this question, no harm intended) Do you dislike RYBKA besically because it's in your eyes, here agreeing with some other lads here, that Rybka is an formally unallowed part of a chain of Fruits? (I'm especially in suspense waiting for your answer on this one.)

6) Your personal belief on the several inventions of the author that shouldnt be named here and his creations? Pro or contra?

7) If a single program is always winning because it's the strongest player, do you support tricks of overflooding a tournament with also good competitors, in special some 'angstgegner', so that the chance that the best could also lose a game or two and thus missing first place? If you understand the tournaments as a social happening for fun? (Please dont misunderstand my questions as if I would insinuate in you always the worst case of several possibilities. I am asking this also to get possible answers from many others. But they might not be so open-minded as you are.)

8) Tord, if you have accomplished to create such a nice program, why dont you like to present it for a knowing community that has much respect for you? If the disturbing noise is a problem for you, couldnt you move this into a corner where all stars in a field place their irritating fan colonies?

9) Could you imagine how fans and collegues want to tear you into the mood of CCT playing with their social communicating via net debates?

Finally I want to thank you in special because you are a nice exception of a programmer talking to a pure lay. I promise that the above is all I had to ask you. Please answer in email what you think is not so possible in public. Again I will keep it for myself and not continue to distract you. So nothing to worry. All the best. Rolf

[bob]

Tord, there is a logical lack or failure in the above. First of all my compliments to you, the author of such a good Glaurung program. I wished you would use it yourself in tournaments.

Which I already do. Just not in the CCT.

What means proprietary?

The definitions of the terms "proprietary" and "free" with respect to computer software have well established meanings. Read through the documents at the FSF website if you are interested.

Are you arguing that if that is forbidden to do that a user or lover of your program could still appear in tournaments as if it were his/her own program? Apparently not!

I can't even parse the question, so I can't give an answer. The same thing applies to the next few questions from your side: I have no idea what you are talking about. I'm sorry.

Where would you disagree? On the new users' side or on the tournament organisers' side?

I see an aspect in all this that is based on a logical fault. People might argue that if it's allowed for Rybka then why not for Glaurung too? Why not for Toga too?

The fault is to compare a commercial program like Rybka, a provable and singular creation by its identified author with all the commercial laws, and many user tweaked versions of free programs like yours for example. If Toga were comparably a singular entity like Rybka then why it's not been sold like Rybka? Apparently because then under commercial laws it would came out that the whole entity isnt singular, at least this is what I suspect.

And finally there is another fallacy resulting out of the typical tradition of computerchess from the times of the wooden hardware chess players. Also then users in thousands would have liked to participate in competitions like owners so to speak. But that worked only for private but not in championships. But it's true that the companies themselves tried to enter their product several times in parallel mode. Perhaps this is now the reason for what we see in discussions like this here.

You argue like a company at the old times which had said that they wouldnt take part, but the users might give it a try. In short why dont you participate yourself?

Finally a question I understand and can answer: I don't play in the CCT because I think it's an awfully dull tournament. I prefer tournaments where I can meet the programmers face to face and go out and have a dinner or a few beers afterwards. Online tournaments are unsatisfactory in general, and CCT even more so, because the ICC is such a noisy place to play.

Because only you had the proprietary right for Glaurung? It cant be a time aspect because what you spend in debates like that isnt less than what you had to invest for a typical CCT event, no?

I spend far less time reading and posting to the CCC over a week than I would have spent in a single weekend playing the CCT. Besides, the amount of time is not the only important thing. I visit CCC briefly several times per day when I want a little break from what I am doing. This is very different from spending several hours Saturday and Sunday night on a chess server. Like most people, I prefer to socialize in real life on Saturday and Sunday night.

Or, if I am totally wrong, again like others would say, why are you so interested in such a topic at all because I never see Fabien in these debates.

Fabien isn't doing computer chess at the moment and doesn't even read this forum, so I don't see why you are surprised not to see any comments from him.

Dont you realise that your open source project which says freedom of use, without your own tournament interest/participation looks strange the moment you then argue and dominate in such discussions? Dont you see this contradiction?

I don't dominate it at all. The only reason I posted at all in this thread was that Bob compared the case of Fruit/Toga to the case of Glaurung/Stockfish and Crafty/some private Crafty clone. I pointed out that the three cases are entirely different, and that the case of Fruit/Toga is the only one which is worth discussing. I would have preferred to leave the discussion there, but other people keep bringing up me and my program again and again, even though it is irrelevant and without interest in the current thread.

Tord

I do not see how you think the cases are different. I am always talking in the context of ICGA events and CCT/ACCA type events. Where the rule has _always_ been "no more than one entry from a single author" which includes derivative works as well. So with Glaurung/Stockfish, you as the original could choose which could participate in the next WCCC and the ICGA would accept that. But only _one_ could enter. Ditto for the CCT events and ACCA events. That is exactly the same issue as the fruit/toga case, and the many past crafty/xyz cases. Your license agreement has _nothing_ to do with this specific issue, nor does mine, nor does Fruit's. This is not about legal use of source code, it is about a specific tournament restriction we have used for _many_ years.

BTW the CCT events are probably more interesting than you give credit for. We've had many a good technical discussion at these events. You will find authors present. Far more than what attends the smallish WCCC and equivalent tournaments. And there's a _whole_ lot less pressure on the operators since they don't do anything, leaving the entire time for game analysis and algorithm discussion. You might be pleasantly surprised if you enter one to see for yourself.

[Tord Romstad]

I don't really feel very comfortable about answering questions about myself and my engine in a thread which really has nothing to do with either, but I'll do it this single time:

1) Is it for you a difference if you would end on place 12 in CCT or someone else as operator/pretended owner of basically your engine?

No. I don't care on what place my engine, nor any other engine, ends up in some tournament. This is partly because I have no interest in the competitive side of computer chess, and partly because I know that the results of a single short tournament have hardly any statistical significance (especially when everyone is running on different hardware). Tournaments are about fun and publicity, not about results.

2) Do you see a similarity in the status of FRUIT and then GLAURUNG?

In what respect? It's an extremely general question.

3) Do you call for as many as possible Glaurung entries in CCT?

No, of course not. There should never be more than one entry of any single engine in the tournament, of course.

4) Do you like to see Toga entry and other Fruit clones?

Personally, I am far more interested in seeing the real Fruit play.

5) (Please excuse the clarity in this question, no harm intended) Do you dislike RYBKA besically because it's in your eyes, here agreeing with some other lads here, that Rybka is an formally unallowed part of a chain of Fruits? (I'm especially in suspense waiting for your answer on this one.)

I think that question has been debated to death a long time ago, and that this thread is not an appropriate place to revive it.

6) Your personal belief on the several inventions of the author that shouldnt be named here and his creations? Pro or contra?

My personal belief about the author (if I guess who you are talking about) and his program is so strongly colored by information I have been told confidentially (and may not remember accurately) during live computer chess tournaments that it would not be appropriate to go into any details on a public forum.

7) If a single program is always winning because it's the strongest player, do you support tricks of overflooding a tournament with also good competitors, in special some 'angstgegner', so that the chance that the best could also lose a game or two and thus missing first place? If you understand the tournaments as a social happening for fun? (Please dont misunderstand my questions as if I would insinuate in you always the worst case of several possibilities. I am asking this also to get possible answers from many others. But they might not be so open-minded as you are.)

I think everybody who is interested in a computer chess tournament, including the author of the presumably strongest program, will always want to see as many strong participants as possible.

8) Tord, if you have accomplished to create such a nice program, why dont you like to present it for a knowing community that has much respect for you? If the disturbing noise is a problem for you, couldnt you move this into a corner where all stars in a field place their irritating fan colonies?

I'm not sure what you mean, but neither I nor my program is hiding. You can download my engine for just about any actively developed operating system, and compile it from source code on many others. It ships with a complete GUI on two different platforms. It's found on most of the major tournaments, and I have operated it myself in tournaments numerous times.

9) Could you imagine how fans and collegues want to tear you into the mood of CCT playing with their social communicating via net debates?

I don't think most of them care much about it.

Tord

[Tord Romstad]

The only reason I posted at all in this thread was that Bob compared the case of Fruit/Toga to the case of Glaurung/Stockfish and Crafty/some private Crafty clone. I pointed out that the three cases are entirely different, and that the case of Fruit/Toga is the only one which is worth discussing. I would have preferred to leave the discussion there, but other people keep bringing up me and my program again and again, even though it is irrelevant and without interest in the current thread.

I do not see how you think the cases are different. I am always talking in the context of ICGA events and CCT/ACCA type events. Where the rule has _always_ been "no more than one entry from a single author" which includes derivative works as well.

Yes, the three cases are similar in the respect that no more than one entry from a single author should ever be allowed to participate. But this is not what we are discussing: It would be pointless to discuss, because everybody agrees about it.

So with Glaurung/Stockfish, you as the original could choose which could participate in the next WCCC and the ICGA would accept that. But only _one_ could enter. Ditto for the CCT events and ACCA events. That is exactly the same issue as the fruit/toga case, and the many past crafty/xyz cases. Your license agreement has _nothing_ to do with this specific issue, nor does mine, nor does Fruit's. This is not about legal use of source code, it is about a specific tournament restriction we have used for _many_ years.

Yes, there is a difference, and it does depend on the license: The difference is which single version of Crafty, Glaurung or Fruit should be allowed to participate. In the case of Crafty, it's obviously your own, official version of Crafty. In the case of Glaurung, it's obviously the first version somebody registers for the tournament. In the case of Fruit, it becomes messy, because it started its life as a free program, but later became proprietary. That's why Fruit is the only case worth discussing.

BTW the CCT events are probably more interesting than you give credit for. We've had many a good technical discussion at these events. You will find authors present.

Well, "interesting" is a very subjective term.

Although I've never been operating my own program during the CCT, I've been logged on and following the games at least in two of the past tournaments. I even recall chatting with you during a Glaurung-Crafty game.

I speak from experience when I say that I just don't find it interesting or enjoyable. There are far worse ways to spend a weekend, but also many better ways. Don't get me wrong: The CCT is a great event, and I am very happy that programmers who lives in parts of the world where there are few computer chess tournaments have a tournament where they can easily participate. But because I'm lucky enough to live in Europe and have plenty of more attractive tournaments to play in, I don't consider the CCT to be worth the time for me.

Far more than what attends the smallish WCCC and equivalent tournaments. And there's a _whole_ lot less pressure on the operators since they don't do anything, leaving the entire time for game analysis and algorithm discussion.

I never feel any sort of pressure in live events either -- perhaps because I don't take computer chess very seriously. It's just a game, and it isn't even I who am playing.

You might be pleasantly surprised if you enter one to see for yourself.

I won't: I've been there, as I said.

Tord

[michiguel]

The only reason I posted at all in this thread was that Bob compared the case of Fruit/Toga to the case of Glaurung/Stockfish and Crafty/some private Crafty clone. I pointed out that the three cases are entirely different, and that the case of Fruit/Toga is the only one which is worth discussing. I would have preferred to leave the discussion there, but other people keep bringing up me and my program again and again, even though it is irrelevant and without interest in the current thread.

I do not see how you think the cases are different. I am always talking in the context of ICGA events and CCT/ACCA type events. Where the rule has _always_ been "no more than one entry from a single author" which includes derivative works as well.

Yes, the three cases are similar in the respect that no more than one entry from a single author should ever be allowed to participate. But this is not what we are discussing: It would be pointless to discuss, because everybody agrees about it.

So with Glaurung/Stockfish, you as the original could choose which could participate in the next WCCC and the ICGA would accept that. But only _one_ could enter. Ditto for the CCT events and ACCA events. That is exactly the same issue as the fruit/toga case, and the many past crafty/xyz cases. Your license agreement has _nothing_ to do with this specific issue, nor does mine, nor does Fruit's. This is not about legal use of source code, it is about a specific tournament restriction we have used for _many_ years.

Yes, there is a difference, and it does depend on the license: The difference is which single version of Crafty, Glaurung or Fruit should be allowed to participate. In the case of Crafty, it's obviously your own, official version of Crafty. In the case of Glaurung, it's obviously the first version somebody registers for the tournament. In the case of Fruit, it becomes messy, because it started its life as a free program, but later became proprietary. That's why Fruit is the only case worth discussing.

That is why, I insist, this should be clearly stated (in writing) in the rules of the tournament. In case there two entries with the same author, the rules should make clear which one has priority. I think that the rules of CCT could be improved a little.

"3a. Each participant (engine) must an original work. No entry can contain code from another program, or be a "clone" of another program. This includes any "personality" settings of an originating program."

"Only the original author, Operator or a team member of the original program may enter and operate."


According to this, only Glaurung can compete, but not Stockfish, even if you allow it. Stockfish is not an original work.

In addition, if we are strict, no program that supports Nalimov EGTBs can enter. "No entry can contain code from another program". I agree with that, but it is not enforced.

In fact, if Glaurung contains code present in Viper, maybe it could not participate either

I know, I know...

Miguel

BTW the CCT events are probably more interesting than you give credit for. We've had many a good technical discussion at these events. You will find authors present.

Well, "interesting" is a very subjective term.

Although I've never been operating my own program during the CCT, I've been logged on and following the games at least in two of the past tournaments. I even recall chatting with you during a Glaurung-Crafty game.

I speak from experience when I say that I just don't find it interesting or enjoyable. There are far worse ways to spend a weekend, but also many better ways. Don't get me wrong: The CCT is a great event, and I am very happy that programmers who lives in parts of the world where there are few computer chess tournaments have a tournament where they can easily participate. But because I'm lucky enough to live in Europe and have plenty of more attractive tournaments to play in, I don't consider the CCT to be worth the time for me.

Far more than what attends the smallish WCCC and equivalent tournaments. And there's a _whole_ lot less pressure on the operators since they don't do anything, leaving the entire time for game analysis and algorithm discussion.

I never feel any sort of pressure in live events either -- perhaps because I don't take computer chess very seriously. It's just a game, and it isn't even I who am playing.

You might be pleasantly surprised if you enter one to see for yourself.

I won't: I've been there, as I said.

Tord

[Rolf]

Don't get me wrong: The CCT is a great event, and I am very happy that programmers who lives in parts of the world where there are few computer chess tournaments have a tournament where they can easily participate. But because I'm lucky enough to live in Europe and have plenty of more attractive tournaments to play in, I don't consider the CCT to be worth the time for me.
Tord

First of all I want to thank you for the answers to my many questions. I'd like to let it all stand like that because also if other questions could be added, I think the way you answered me is also answer itself. Thanks again for your support.

To the above I have a little comment. Honestly I see a big contradiction. If I can rightfully assume that you are a dedicated computerchess programs lover (time and motivation height) and if that is true it's absolutely strange that you intentionally want to skip the CCT as a better alternative - since the international articipation - to the poor events like the last one in China with the unacceptable, intolerable split into VIP functioneers and the brainers. Just take a look at the photos from the last event in Alabama UAB. Why do you need beer or whatever, if the whole essence of the game is in your mind? Do you need live webcam reports? That might be do-able.

You are really cruel the way you tell Bob who has no chance to attend live tournements in Europe, that you are living in Europe and that you therefore dont need the CCT! Please consider this. Just do it out of thankfulness for the things you learned via net debates with international people. Unthinkable that in soccer Germany, even Norway, would declare abstinence because of this or that and the bad weather. You are part of the community, so let it roll! Also think about the many programmers who will participate for the first time.

[bob]

Don't get me wrong: The CCT is a great event, and I am very happy that programmers who lives in parts of the world where there are few computer chess tournaments have a tournament where they can easily participate. But because I'm lucky enough to live in Europe and have plenty of more attractive tournaments to play in, I don't consider the CCT to be worth the time for me.
Tord

First of all I want to thank you for the answers to my many questions. I'd like to let it all stand like that because also if other questions could be added, I think the way you answered me is also answer itself. Thanks again for your support.

To the above I have a little comment. Honestly I see a big contradiction. If I can rightfully assume that you are a dedicated computerchess programs lover (time and motivation height) and if that is true it's absolutely strange that you intentionally want to skip the CCT as a better alternative - since the international articipation - to the poor events like the last one in China with the unacceptable, intolerable split into VIP functioneers and the brainers. Just take a look at the photos from the last event in Alabama UAB. Why do you need beer or whatever, if the whole essence of the game is in your mind? Do you need live webcam reports? That might be do-able.

You are really cruel the way you tell Bob who has no chance to attend live tournements in Europe, that you are living in Europe and that you therefore dont need the CCT! Please consider this. Just do it out of thankfulness for the things you learned via net debates with international people. Unthinkable that in soccer Germany, even Norway, would declare abstinence because of this or that and the bad weather. You are part of the community, so let it roll! Also think about the many programmers who will participate for the first time.

Rolf, he is not trying to be "cruel" to Bob or anyone else. The nice thing about computer chess events is that they are _all_ optional. NO attendance required. Which makes them fun, in and of themselves. I do not try to insult those in Europe that can attend the local events, when I don't go. I don't assume they have some sort of dark motive when they choose to not come over here. And if someone really is not interested in net-wise tournaments at all, that's their choice and they are free to exercise it with no strings or stigma attached...

I'd love to see _everyone_ participate in the CCT events. They have been a blast, and while I agree that they are not as much fun as the old ACM events I attended every year for 20 years, they are still a lot of fun and get the mental juices flowing each time one comes around...

--from someone that has probably done way too many of any kind of computer chess event...

[Ryan Benitez]

I'll add that Toga is based on Fruit 2.2.1, which is vastly different from Fruit 2.3.1 and subsequent versions (according to Ryan).

It has been a long time so I do not remember what changes happened at what version but Fruit has a few big changes in that it is a bitboard engine. I do not know if any of the Togas have made such a conversion yet. Naturally Fruit has a new eval to go along with that. It was harder to make a new eval that even matched the fast and simple old one. If the goal was just elo gain it would not have been worth the time compared to other areas to work on.

If the Toga developers haven't contributed a single idea, then why is the latest Toga almost 100 elo stronger than Fruit 2.2.1? That seems to paint Fabien as inept (which he is far from being) because some amateurs could rewrite a few lines of code and get 100 elo improvement.
If a commercial programmer could get that much improvement, they'd be selling us a new engine. In fact, some do so with far less a gain.

Not every programmer or team has the same goals. My guess for the Thomas version of Toga is that the goal is ELO gain with as few changed lines as possible. The obvious way to go if that is your goal is to focus on tuning search pruning. To some programmers such tuning may be fun but to others this may be a tedius task leading them to work on other parts of the program more entertaining to them. If you need an example of how much gain can be had by tuning an already good program look at the gain from Fruit 2.2.1 to Fruit051103. I can trim the lines of code difference between the 2 version to less than 50 lines of code. The gain is in the tuning and thanks to a great testing team. If I was a commercial chess programmer with the goal of making money I would be begging the old Fruit beta testers to join my team.

[Uri Blass]

I'll add that Toga is based on Fruit 2.2.1, which is vastly different from Fruit 2.3.1 and subsequent versions (according to Ryan).

It has been a long time so I do not remember what changes happened at what version but Fruit has a few big changes in that it is a bitboard engine. I do not know if any of the Togas have made such a conversion yet. Naturally Fruit has a new eval to go along with that. It was harder to make a new eval that even matched the fast and simple old one. If the goal was just elo gain it would not have been worth the time compared to other areas to work on.

I wonder what parts of the playing engine are the same as fruit in the case of bitboard engine.

move generator is clearly different.

The evaluation needs a new code and I understand that you did not try to translate the old evaluation to bitboard but made something new.

Is it correct that practically all the similiarity is in the search algorithm?

Uri

[Ryan Benitez]

I'll add that Toga is based on Fruit 2.2.1, which is vastly different from Fruit 2.3.1 and subsequent versions (according to Ryan).

It has been a long time so I do not remember what changes happened at what version but Fruit has a few big changes in that it is a bitboard engine. I do not know if any of the Togas have made such a conversion yet. Naturally Fruit has a new eval to go along with that. It was harder to make a new eval that even matched the fast and simple old one. If the goal was just elo gain it would not have been worth the time compared to other areas to work on.

I wonder what parts of the playing engine are the same as fruit in the case of bitboard engine.

move generator is clearly different.

The evaluation needs a new code and I understand that you did not try to translate the old evaluation to bitboard but made something new.

Is it correct that practically all the similiarity is in the search algorithm?

Uri

Yes, very little in the search has changed from Fruit051103 to the last version. The 2 changes from Fruit051103 are to null move and tuning by the beta testers to LMR.