Please stop developing free Stockfish

[Albert Silver]

Everyone by now knows that Stockfish is being developed for free.
I have a plan to get the authors out of this "Free Stockfish Problem".

5. Once a considerable elo has been gained, offer that version for sale for $75 to a $100 or whatever the market will pay.
6. The original authors can keep 50% of the profits and split the rest to the others that contributed to the improvement.

kgburcham

Implemented:



SilvianR

That is pretty expensive source code.

[ZirconiumX]

Everyone by now knows that Stockfish is being developed for free.
I have a plan to get the authors out of this "Free Stockfish Problem".

5. Once a considerable elo has been gained, offer that version for sale for $75 to a $100 or whatever the market will pay.
6. The original authors can keep 50% of the profits and split the rest to the others that contributed to the improvement.

kgburcham

Implemented:



SilvianR

That is pretty expensive source code.

Marco has a family to feed, you know!

Matthew:out

[Henk]

It may be possible to stop further improvement of Stockfish. Just offer the programmers that work on Stockfish some money if they stop their developments.

On the other hand a programmer that has a good idea that only works for Stockfish, with a test that gives an improvement of say 50 ELO may go to a concurrent and say I will not implement my idea in Stockfish if you pay me some money. (Looks a bit like blackmail to me).

[Evert]

It may be possible to stop further improvement of Stockfish. Just offer the programmers that work on Stockfish some money if they stop their developments.

Fantastic idea. I don't know about the SF team, but I for one will promise that I will not try to improve Stockfish if you pay me a modal income for the next ten years.

Deal?

[Henk]

It may be possible to stop further improvement of Stockfish. Just offer the programmers that work on Stockfish some money if they stop their developments.

Fantastic idea. I don't know about the SF team, but I for one will promise that I will not try to improve Stockfish if you pay me a modal income for the next ten years.

Deal?

You have to go to Houdini team or the Komodo team I guess. My chess program plays about 1000 ELO worse than Stockfish and so won't sell at all.
So there is no gain for me. I don't have Komodo and Houdini shares too.

[Uri Blass]

It may be possible to stop further improvement of Stockfish. Just offer the programmers that work on Stockfish some money if they stop their developments.

On the other hand a programmer that has a good idea that only works for Stockfish, with a test that gives an improvement of say 50 ELO may go to a concurrent and say I will not implement my idea in Stockfish if you pay me some money. (Looks a bit like blackmail to me).

No need to offer money to convince other not to work for improvement of stockfish.

Some people do it well enough by complaining about wasting of the framework time.

It clearly convinced me not to send patches to try to improve stockfish so nobody in the future is going to blame me that I waste the framework time.

I am responsible only for a small part of the improvement of stockfish but
I strongly believe that the improvement of stockfish is going to be slightly slower without my patches but it is obvious that getting progress as fast as possible is not what the stockfish team prefers.

My opinion is that if the target is to get improvement as fast as possible then the best is simply to allow every test at least in low priority even if it is obvious that the target of it is not to help stockfish to become better.

I think that the stockfish team simply miss the psychological point that if you allow people to test everything they want in the framework in a low priority then it can encourage them also to give patches to improve stockfish with higher priority.

The problem of people who are not interested in testing useless stuff can be solved easily because it is possible to allow testers not to test patches with low priority(and my guess is that part of the people who give computer time have no objection to test everything because they also understand the psychological point that I mention so they think that allowing testing everything can help stockfish to become better indirectly).

[Henk]

Maybe these people who slow down SF development are already paid to do so. If not a concurrent commercial party should pay these members to keep doing their good work. (Sorry just kidding)

[Uri Blass]

Maybe these people who slow down SF development are already paid to do so. If not a concurrent commercial party should pay these members to keep doing their good work.

There is no money or almost no money in computer chess today so I do not believe that people are payed to slow down the SF developement.

[Henk]

Maybe these people who slow down SF development are already paid to do so. If not a concurrent commercial party should pay these members to keep doing their good work.

There is no money or almost no money in computer chess today so I do not believe that people are payed to slow down the SF developement.

Is that because of the availability of free strong playing chess programs ? Or because people lost interest in computer chess ?

[rlsuth]

There is no money or almost no money in computer chess today so I do not believe that people are payed to slow down the SF developement.

Is that because of the availability of free strong playing chess programs ? Or because people lost interest in computer chess ?

It is more to do with the strength of the top programs. The field has pretty much been conquered now. Stockfish and Houdini are so strong that it really doesn't matter if they gain 50 ELO or not. There is no incentive for new programmers to develop new chess engines because, unless your program magically surpasses the strongest engine immediately, nobody is going to pay for it.

The drop in interest in chess programs began when Kasparov was beaten and Grandmasters stopped playing computers. It can be seen in the demise of the computer chess magazines and dedicated chess machines.

I would absolutely buy a nice wood dedicated chess machine board if one were available today, with a decent engine and output to a computer monitor. It should be possible to create a really great one using a NUC and a modified Houdini. If the author also provided an interface to ICC/ Playchess so you could use the board to play interent chess I think there would be a large audience, especially if it also included the computer moving it's own pieces, maybe magnetically? How about it Robert?