Fat Fritz 2.0

ChessBase has published Fat Fritz 2.0, on the base of SF dev it seems, with a new, larger net. An article with some explanations and marketing is here: https://en.chessbase.com/post/fat-fritz-2-0-the-new-number-1

Will be interesting if CB will release the (modified) source code of SF.

It seems it has been done, Ed. There were already some comments on talkchess about these sources, but I am not an expert, as you know well. There seems to be just a few changes. The project is similar to FF1, where SF has replaced LC0, with a new net mainly.

Here is a link given by MikeB: https://github.com/DanielUranga/Stockfish/tree/ff2

matejst wrote:

ChessBase has published Fat Fritz 2.0, on the base of SF dev it seems, with a new, larger net. An article with some explanations and marketing is here: https://en.chessbase.com/post/fat-fritz-2-0-the-new-number-1

I would not recommend buying FF 2.0. It is basically Stockfish with modifications to run bigger nets(40mb). This could be good, but current results show it is still inferior to the best Stockfish versions running the 20mb nets.

If you want the Fritz gui and the the perks that come with buying FF 2.0. Then buy FF 2.0 version over Fritz 17 version of the GUI.

But for 99 E  just for FF 2.0 is a waste of money.

And we can be pretty sure that if running bigger nets is better for stockfish. It will be added without cost in the official version of Stockfish.

mwyoung wrote:

matejst wrote:

ChessBase has published Fat Fritz 2.0, on the base of SF dev it seems, with a new, larger net. An article with some explanations and marketing is here: https://en.chessbase.com/post/fat-fritz-2-0-the-new-number-1

I would not recommend buying FF 2.0. It is basically Stockfish with modifications to run bigger nets(40mb). This could be good, but current results show it is still inferior to the best Stockfish versions running the 20mb nets.

If you want the Fritz gui and the the perks that come with buying FF 2.0. Then buy FF 2.0 version over Fritz 17 version of the GUI.

But for 99 E  just for FF 2.0 is a waste of money.

And we can be pretty sure that if running bigger nets is better for stockfish. It will be added without cost in the official version of Stockfish.

FF and ChessBase could be totally destroyed by a very high quality community built GUI with all bells and whistles, plug and play, plus absolute ease of inserting any engine you like, or already coming with engines and nets.
Comp chess community is pretty crap at designing good GUIs for ease of use by chess players, but such a product would well answer these commercial spin-off exploitations.

matejst wrote:

It seems it has been done, Ed. There were already some comments on talkchess about these sources, but I am not an expert, as you know well. There seems to be just a few changes. The project is similar to FF1, where SF has replaced LC0, with a new net mainly.

Here is a link given by MikeB: https://github.com/DanielUranga/Stockfish/tree/ff2

First Lc0, now SF12.

Glad I am out of business.

The double net is just a lightning rod.

Chris Whittington wrote:

FF and ChessBase could be totally destroyed by a very high quality community built GUI with all bells and whistles, plug and play, plus absolute ease of inserting any engine you like, or already coming with engines and nets.
Comp chess community is pretty crap at designing good GUIs for ease of use by chess players, but such a product would well answer these commercial spin-off exploitations.

I completely agree. The Banksia GUI had a lot of potential, but it went another way, and became almost unusable for anything else but eng-eng matches and tournaments.

matejst wrote:

Chris Whittington wrote:

FF and ChessBase could be totally destroyed by a very high quality community built GUI with all bells and whistles, plug and play, plus absolute ease of inserting any engine you like, or already coming with engines and nets.
Comp chess community is pretty crap at designing good GUIs for ease of use by chess players, but such a product would well answer these commercial spin-off exploitations.

I completely agree. The Banksia GUI had a lot of potential, but it went another way, and became almost unusable for anything else but eng-eng matches and tournaments.

The problem is simple, it’s not ChessBase ripping off open source, because they are free to publish like however they do. The problem is that Stockfish and co are NOT fully finished products so CB or anyone can step in, finish them, and publish. To prevent CB adding value and making money, well, turn SF and co into finished product, suitable for users.
Emulate the FF interface, do what it can do, including all manner of compatibility with other chessBase products and data, and bundle it with all top engines (GPL?) with ease of download. That should do the trick. Everything open source.

Don’t let any long-time academics anywhere near it.