Fruit vs. Toga poll

[michiguel]

A couple of hours ago there was a poll about their participation in CCT. I was going to comment about it. I came back and the thread & poll is gone. What happened?

Miguel

[Eelco de Groot]

Somebody felt the need to post a link to his favourite pirate software as the next CCT 2009 candidates. Poll vanished into moderators subforum.

My comment was

I voted for yes, provided it does not endanger the participation of an entry by Ryan Benitez or Fabien Letouzey or another member of the Fruit team that would be the official Fruit representative, provided also that Fabien and Ryan don't object to Toga participating, - I can't imagine that Thomas Gaksch would object-, and it would be an official Toga entry by the TDDB forum, because I see that as the best way to limit the number of Toga entries to one actively developed fork of Fruit, a community effort from the TDDB programmers.

but when I wanted to post this Christopher's sabotage had succeeded and the poll from Peter Skinner had disappeared already...

[Spock]

Indeed, my reply was all ready as well but when I hit the submit button the thread was gone. Here was my comment:

I fully appreciate how programmers that have spent hundreds or thousands of hours creating their own programmes from scratch feel, but in my view

- Toga is stronger than the last public versions of Fruit
- it has probably diverged quite a bit from Fruit by now
- it abides by the GPL and is 100% legal
- open source is a good thing and should be supported

So I'd say let it participate. Even if Fruit was there as well.

[CThinker]

I too tried to reply to the poll when it disappeared.

Anyhow, here is my comment on the subject:

The difference between Thinker version 4.x and version 5.x is very big. A rewrite of 95% of code. The difference between 'active', 'passive' and 'inert' version of Thinker 5.x are at least 30% (the entire eval code of each, some sections of search and initializations).

The same can be said with Crafty 19.x and 20.x. This is massive rewrite by Bob.

Even between major versions, Thinker 5.4 is dramatically different from Thinker 5.0.

Those differences dwarf the differences between Fruit and its clones, or between Toga and its clones.

Should I then be allowed to enter 20 versions of Thinker in CCT, and Bob to enter another 20 versions of Crafty?

Every new version of known engines out there is a change from the previous version. Most of the time, those changes are big. And yet, we don't think of them as new engines. Bob changed thousands of lines of code between Crafty 19.x and 20.x. But it is still Crafty. Now, someone changes less than a hundred lines of code of Fruit, then all of a sudden its a "new" engine?

[bob]

Somebody felt the need to post a link to his favourite pirate software as the next CCT 2009 candidates. Poll vanished into moderators subforum.

My comment was

I voted for yes, provided it does not endanger the participation of an entry by Ryan Benitez or Fabien Letouzey or another member of the Fruit team that would be the official Fruit representative, provided also that Fabien and Ryan don't object to Toga participating, - I can't imagine that Thomas Gaksch would object-, and it would be an official Toga entry by the TDDB forum, because I see that as the best way to limit the number of Toga entries to one actively developed fork of Fruit, a community effort from the TDDB programmers.

but when I wanted to post this Christopher's sabotage had succeeded and the poll from Peter Skinner had disappeared already...

My vote is simple. "NO". Derivative works are not allowed. I'd prefer to see Toga as based on my cluster testing it is stronger than Fruit. But as far as I am concerned, it is an "either/or" type decision. Otherwise several have modified versions of Crafty they would like to enter, and that makes no sense IMHO. Not to mention glaurung/stockfish, and so forth.

[Harvey Williamson]

If Fruit is entered then Toga should not be. If Fruit is not entered then there is a debate to be had but my feeling is still no.

[Tord Romstad]

My vote is simple. "NO". Derivative works are not allowed. I'd prefer to see Toga as based on my cluster testing it is stronger than Fruit. But as far as I am concerned, it is an "either/or" type decision. Otherwise several have modified versions of Crafty they would like to enter, and that makes no sense IMHO. Not to mention glaurung/stockfish, and so forth.

I think the three cases are fundamentally different:

• Crafty's license states very clearly that it is not allowed for anyone to enter the program in any tournament without the author's explicit permission. Clearly, this implies that the "official" Crafty is the only one which can participate in a tournament like CCT.
• For Glaurung/Stockfish, the situation is precisely the opposite: Both programs are free, which means that anyone is allowed to enter either of these programs in the CCT, without asking the author(s) for permission (because this permission is already included in the license). Obviously, both programs cannot participate in the same tournament if the tournament rules prohibit multiple programs sharing code. If someone enters Stockfish before someone enters Glaurung, only Stockfish can participate, and vice versa.
• For Fruit/Toga, the situation is a mess, because they are based on the same code based, and one is proprietary and the other is free. I have no idea what the rules should be for this strange situation, and I wouldn't have voted in the poll if it was still visible. It's an interesting question, unlike the cases with Crafty and Stockfish/Glaurung, where everything seems crystal clear.

Tord

[michiguel]

My vote is simple. "NO". Derivative works are not allowed. I'd prefer to see Toga as based on my cluster testing it is stronger than Fruit. But as far as I am concerned, it is an "either/or" type decision. Otherwise several have modified versions of Crafty they would like to enter, and that makes no sense IMHO. Not to mention glaurung/stockfish, and so forth.

I think the three cases are fundamentally different:

• Crafty's license states very clearly that it is not allowed for anyone to enter the program in any tournament without the author's explicit permission. Clearly, this implies that the "official" Crafty is the only one which can participate in a tournament like CCT.
• For Glaurung/Stockfish, the situation is precisely the opposite: Both programs are free, which means that anyone is allowed to enter either of these programs in the CCT, without asking the author(s) for permission (because this permission is already included in the license). Obviously, both programs cannot participate in the same tournament if the tournament rules prohibit multiple programs sharing code. If someone enters Stockfish before someone enters Glaurung, only Stockfish can participate, and vice versa.
• For Fruit/Toga, the situation is a mess, because they are based on the same code based, and one is proprietary and the other is free. I have no idea what the rules should be for this strange situation, and I wouldn't have voted in the poll if it was still visible. It's an interesting question, unlike the cases with Crafty and Stockfish/Glaurung, where everything seems crystal clear.

Tord

I think it is simple. If the rules of the tournament says that only one entry per author is allowed, then either Toga or Fruit is allowed, not both.

The only debate will be which one: if it is determined by earlier registration or by authorization of the common author (Fabien).

Miguel

[Tord Romstad]

My vote is simple. "NO". Derivative works are not allowed. I'd prefer to see Toga as based on my cluster testing it is stronger than Fruit. But as far as I am concerned, it is an "either/or" type decision. Otherwise several have modified versions of Crafty they would like to enter, and that makes no sense IMHO. Not to mention glaurung/stockfish, and so forth.

I think the three cases are fundamentally different:

• Crafty's license states very clearly that it is not allowed for anyone to enter the program in any tournament without the author's explicit permission. Clearly, this implies that the "official" Crafty is the only one which can participate in a tournament like CCT.
• For Glaurung/Stockfish, the situation is precisely the opposite: Both programs are free, which means that anyone is allowed to enter either of these programs in the CCT, without asking the author(s) for permission (because this permission is already included in the license). Obviously, both programs cannot participate in the same tournament if the tournament rules prohibit multiple programs sharing code. If someone enters Stockfish before someone enters Glaurung, only Stockfish can participate, and vice versa.
• For Fruit/Toga, the situation is a mess, because they are based on the same code based, and one is proprietary and the other is free. I have no idea what the rules should be for this strange situation, and I wouldn't have voted in the poll if it was still visible. It's an interesting question, unlike the cases with Crafty and Stockfish/Glaurung, where everything seems crystal clear.

Tord

I think it is simple. If the rules of the tournament says that only one entry per author is allowed, then either Toga or Fruit is allowed, not both.

The only debate will be which one: if it is determined by earlier registration or by authorization of the common author (Fabien).

Yes, that's exactly what I meant. In the case of Crafty, it is clear that Bob is the only one who can enter. In the case of Stockfish/Glaurung, it is clear that whoever registers either of the two programs earlier than the other can enter. In the case of Fruit/Toga, I have no idea, but it's an interesting discussion.

Tord

[Steve B]

A couple of hours ago there was a poll about their participation in CCT. I was going to comment about it. I came back and the thread & poll is gone. What happened?

Miguel

Peter Skinner requested that the moderators remove his thread which we then did
Regards
Steve