Fat Fritz 2

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

fohristiwhirl
Posts: 30
Joined: Mon May 11, 2020 11:52 am
Full name: Allan Thomas

Re: Fat Fritz 2

Post by fohristiwhirl »

Ideally we would need a contemporary SF dev version on that list for comparison.
gaard
Posts: 463
Joined: Mon Jun 07, 2010 3:13 am
Location: Holland, MI
Full name: Martin W

Re: Fat Fritz 2

Post by gaard »

Graham Banks wrote: Sat Feb 13, 2021 11:40 pm Latest from the blitz list.

Image
That doesn't speak very well of CCRL. Why is listed as a "Commercial" engine and not "Open source" and a derivative of SF, which it obviously is??
User avatar
Graham Banks
Posts: 44804
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Fat Fritz 2

Post by Graham Banks »

fohristiwhirl wrote: Sat Feb 13, 2021 11:53 pm Ideally we would need a contemporary SF dev version on that list for comparison.
I can't speak for Sergio, but I would imagine that might be on his list.
gbanksnz at gmail.com
connor_mcmonigle
Posts: 544
Joined: Sun Sep 06, 2020 4:40 am
Full name: Connor McMonigle

Re: Fat Fritz 2

Post by connor_mcmonigle »

gaard wrote: Sat Feb 13, 2021 11:37 pm
connor_mcmonigle wrote: Sat Feb 13, 2021 11:31 pm
gaard wrote: Sat Feb 13, 2021 11:20 pm
By separating the network weights from the provided binary, there is no longer a GPL violation.
Not quite. If the NN doesn't allow you to build an executable that is functionally identical to what is being distributed by ChessBase, then the GPL violation remains. Sleight-of-hand makes their case look worse.

Yes. This sleight of hand doesn't look good ethically, but, of course, none of this does in the first place. By separating the weights from the executable, ChessBase is now selling two separate items when one purchases FF2: the weights and the executable. The conveyed executable can be, now, exactly reproduced by compiling the provided GitHub source and, therefore, future sales of FF2 won't be a GPL violation in my understanding. Previous sales of the integrated binary were likely a GPL violation though.
I don't follow. If the publicly available NN is not equivalent to the one distributed by CB then how can you exactly reproduce what is distributed by CB? Or, are you under the assumption that the publicly available NN is equivalent to the embedded one that CB is selling?
ChessBase now is distributing both the weights and the slightly modified Stockfish executable as separate files to customers. The executable is GPL licensed and can be exactly reproduced by compiling the GitHub sources (therefore no violation there). The commercial weights file is not/cannot (as it is data) be GPL licensed. In fact, whether the weights file can be licensed at all is an open question, though ChessBase might try to take some legal action were you to start distributing their commercial weights file without their permission.

When ChessBase was distributing both the commercial weights file and the modified Stockfish executable as a single, integrated executable to customers, ChessBase's failure to provide customers with the weights file was/is a violation as it is impossible to reproduce the GPL licensed, integrated, binary with just the provided GitHub source.

Hopefully this explains why I think ChessBase's selling the executable and weights separately doesn't violate Stockfish's license, while their selling the integrated binary was/is a violation.
Or, are you under the assumption that the publicly available NN is equivalent to the embedded one that CB is selling?
No...
gaard
Posts: 463
Joined: Mon Jun 07, 2010 3:13 am
Location: Holland, MI
Full name: Martin W

Re: Fat Fritz 2

Post by gaard »

connor_mcmonigle wrote: Sun Feb 14, 2021 12:00 am
gaard wrote: Sat Feb 13, 2021 11:37 pm
connor_mcmonigle wrote: Sat Feb 13, 2021 11:31 pm
gaard wrote: Sat Feb 13, 2021 11:20 pm
By separating the network weights from the provided binary, there is no longer a GPL violation.
Not quite. If the NN doesn't allow you to build an executable that is functionally identical to what is being distributed by ChessBase, then the GPL violation remains. Sleight-of-hand makes their case look worse.

Yes. This sleight of hand doesn't look good ethically, but, of course, none of this does in the first place. By separating the weights from the executable, ChessBase is now selling two separate items when one purchases FF2: the weights and the executable. The conveyed executable can be, now, exactly reproduced by compiling the provided GitHub source and, therefore, future sales of FF2 won't be a GPL violation in my understanding. Previous sales of the integrated binary were likely a GPL violation though.
I don't follow. If the publicly available NN is not equivalent to the one distributed by CB then how can you exactly reproduce what is distributed by CB? Or, are you under the assumption that the publicly available NN is equivalent to the embedded one that CB is selling?
ChessBase now is distributing both the weights and the slightly modified Stockfish executable as separate files to customers. The executable is GPL licensed and can be exactly reproduced by compiling the GitHub sources (therefore no violation there). The commercial weights file is not/cannot (as it is data) be GPL licensed. In fact, whether the weights file can be licensed at all is an open question, though ChessBase might try to take some legal action were you to start distributing their commercial weights file without their permission.

When ChessBase was distributing both the commercial weights file and the modified Stockfish executable as a single, integrated executable to customers, ChessBase's failure to provide customers with the weights file was/is a violation as it is impossible to reproduce the GPL licensed, integrated, binary with just the provided GitHub source.

Hopefully this explains why I think ChessBase's selling the executable and weights separately doesn't violate Stockfish's license, while their selling the integrated binary was/is a violation.
Or, are you under the assumption that the publicly available NN is equivalent to the embedded one that CB is selling?
No...
Can you confirm that what they are selling and distributing includes the publicly available NN file in the GitHub repo and not another NN? If so, great. If not, then the infringement remains and nothing has changed. That was the only point I wished to make. If we find contradiction elsewhere than I apologize in advance because that was my only interest.
connor_mcmonigle
Posts: 544
Joined: Sun Sep 06, 2020 4:40 am
Full name: Connor McMonigle

Re: Fat Fritz 2

Post by connor_mcmonigle »

gaard wrote: Sun Feb 14, 2021 12:06 am Can you confirm that what they are selling and distributing includes the publicly available NN file in the GitHub repo and not another NN? If so, great. If not, then the infringement remains and nothing has changed.
Perhaps I wasn't clear in the above. Do you mind reading my previous post again to make sure you're not missing something?
(The opposite has been confirmed, the weights files don't match FYI)
gaard
Posts: 463
Joined: Mon Jun 07, 2010 3:13 am
Location: Holland, MI
Full name: Martin W

Re: Fat Fritz 2

Post by gaard »

connor_mcmonigle wrote: Sun Feb 14, 2021 12:10 am
gaard wrote: Sun Feb 14, 2021 12:06 am Can you confirm that what they are selling and distributing includes the publicly available NN file in the GitHub repo and not another NN? If so, great. If not, then the infringement remains and nothing has changed.
Perhaps I wasn't clear in the above. Do you mind reading my previous post again to make sure you're not missing something?
(The opposite has been confirmed, the weights files don't match FYI)
The misunderstanding then is my own. Thank you for confirming.
Dann Corbit
Posts: 12799
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Fat Fritz 2

Post by Dann Corbit »

gaard wrote: Sat Feb 13, 2021 11:53 pm
Graham Banks wrote: Sat Feb 13, 2021 11:40 pm Latest from the blitz list.

Image
That doesn't speak very well of CCRL. Why is listed as a "Commercial" engine and not "Open source" and a derivative of SF, which it obviously is??
I don't think they have a designation for derivative.
It is both open source and commercial. They don't have a way to paint the names with two different colors, so far as I know.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
gaard
Posts: 463
Joined: Mon Jun 07, 2010 3:13 am
Location: Holland, MI
Full name: Martin W

Re: Fat Fritz 2

Post by gaard »

Dann Corbit wrote: Sun Feb 14, 2021 12:18 am
gaard wrote: Sat Feb 13, 2021 11:53 pm
Graham Banks wrote: Sat Feb 13, 2021 11:40 pm Latest from the blitz list.

Image
That doesn't speak very well of CCRL. Why is listed as a "Commercial" engine and not "Open source" and a derivative of SF, which it obviously is??
I don't think they have a designation for derivative.
It is both open source and commercial. They don't have a way to paint the names with two different colors, so far as I know.
Fair enough. Although, they used to have a designation for legally dubious software, if I'm not mistaken.

At the very least, it is a SF derivative, and seeing something like "This is one of the 4 Fat Fritz versions we tested: Compare them!" is misleading.
carldaman
Posts: 2287
Joined: Sat Jun 02, 2012 2:13 am

Re: Fat Fritz 2

Post by carldaman »

Dann Corbit wrote: Sun Feb 14, 2021 12:18 am
gaard wrote: Sat Feb 13, 2021 11:53 pm
Graham Banks wrote: Sat Feb 13, 2021 11:40 pm Latest from the blitz list.

Image
That doesn't speak very well of CCRL. Why is listed as a "Commercial" engine and not "Open source" and a derivative of SF, which it obviously is??
I don't think they have a designation for derivative.
It is both open source and commercial. They don't have a way to paint the names with two different colors, so far as I know.
What color does Orange and Blue make?
:roll:
I suggest that new color be used to designate FF2. :P