Hacker News new | past | comments | ask | show | jobs | submit login
Stockfish vs. ChessBase, Round 1 (lichess.org)
93 points by rococode on July 23, 2022 | hide | past | favorite | 7 comments



Is there any steel man for what ChessBase is doing here? Even their argument at the end of the article that perhaps they should be punished, but it wouldn’t be fair not to allow them to continue to monetise stockfish seems pretty weak.


Agreed.

Also, where is the src for the original release of Fat Fritz 2? They released a cut down version which they then released partial src for, but irrespective, they still distributed GPL code so owe the recipients the full src.

While I hope chessbase lose this badly, losing all rights to distribute stockfish code in any form, I still think they should also be forced to comply with the license they chose to use code from & release the full src.


They’re just boasting excessively and failing to share source code that probably wouldn’t actually contribute anything.

Yea they should be behaving differently but both parties are being a bit childish. Settle for an apology and a code release and be done with it.


I fail to see how Stockfish is being childish here. Sounds like ChessBase was given ample opportunity to comply (release the actual source) and tried to weasel out of it (by releasing a modified copy of their new engine's source which hides the secret sauce). Even so, Stockfish only tried to revoke the license when they came to find out ChessBase had ignored the GPL requirements of Stockfish with not one, but two of the engines they sell.

How far should you let someone fuck around with your license terms before you decide to exercise the nuclear option?


> How far should you let someone fuck around with your license terms before you decide to exercise the nuclear option?

Game theory (specifically the iterated prisoners' dilemma) says: once. Many results from psychology - especially parenting - support the same conclusion. Both punishment/retribution for bad behavior and reward for improving behavior should be swift and unambiguous. It's ChessBase's turn to cooperate and show some good faith. Words alone won't do it.


This isnt at all related to rhe original topic but if you like game theory, you'll love this paper[1]. (Tl;dr tit-for-tat isn't the optimal strategy after all).

Also, in repeated prisoner's dilemma with random errors, it has been long known that tit-for-tat isn't the best strategy either.

[1]: http://www.pnas.org/content/109/26/10409.full


> Settle for an apology and a code release and be done with it

My understanding is that ChessBase previously "released the code", and by "released the code" I mean released an intentionally neutered version that didn't include all the code that was still not in compliance with the license.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: