Trusted testers wanted

[mcostalba]

I don't see why this is so hard to understand. If you take a freely distributed program and modify it, you have two options.

(1) keep it to yourself. You can not distribute it to another person or group, period. You can use it solely for your own purposes.

(2) Distribute it to the world, just as the original program you _copied_ was distributed.

It really is that simple, and that specific, and that clear.

There is a third possibility

(3) keep it to yourself and your codevelopers _while_ you are still developing / testing. Then, when you release publicy you have to release togheter with the source code (or the source code should be provided upon request)

During development of last version of Stockfish I have kept git repository private (it still is), only Joona and Tord had a copy of it.

This is common also in other GPL developments where code is released according to GPL only when it is finished. Not all GPL projects works like this, as example Linux kernel and git itself they are fully open, it means they have their git repositories publicly available and everybody can follow the development patch by patch. Stockfish instead is fully disclosed togheter with the sources, well, actually just the sources only at the end of each release cycle.

Perhaps one day I will make git repository public, but in this chess engines world, at this moment in time, I think this would be, as we say in Italy "trying to take a step longer then your leg"

[MattieShoes]

Are y'all arguing because you want the code or just for the sake of arguing? I don't see why everything needs to turn into a legal battle. It's an admitted clone, within 10 points of the original, and the original author knows and is okay with what he's doing presently. He's not being secretive, he just wants help testing things. So why all the huffing and puffing? Take a heart pill!

[mcostalba]

Are y'all arguing because you want the code or just for the sake of arguing? I don't see why everything needs to turn into a legal battle. It's an admitted clone, within 10 points of the original, and the original author knows and is okay with what he's doing presently. He's not being secretive, he just wants help testing things. So why all the huffing and puffing? Take a heart pill!

GPL is not in the DNA of chess engine developers. It is admittely an odd thing in an environment where competition and "secret recipes" are the norm, or at least have been the norm in the past.

I think any discussion that involves GPL in chess engines is useful to spread the knowledge, to make chess engine people more conscious of this possibility.

You cannot just say: "It is like this, stop"

It doesn't work like this. To change their minds people needs to talk, talk and again talk, this is a very long process and takes long time.

So I think this kind of discussions are a useful thing to do, especially in the long term.

[Gian-Carlo Pascutto]

Are y'all arguing because you want the code or just for the sake of arguing? I don't see why everything needs to turn into a legal battle. It's an admitted clone, within 10 points of the original, and the original author knows and is okay with what he's doing presently. He's not being secretive, he just wants help testing things. So why all the huffing and puffing? Take a heart pill!

It's a legal battle because what he is doing might be illegal. Very simple... Clone, no clone, strength is completely irrelevant.

[Tord Romstad]

It's a legal battle because what he is doing might be illegal. Very simple... Clone, no clone, strength is completely irrelevant.

It is obviously completely legal. The GPL FAQ even says explicitly that it is:

The GPL does not require you to release your modified version, or any part of it. You are free to make modifications and use them privately, without ever releasing them. This applies to organizations (including companies), too; an organization can make a modified version and use it internally without ever releasing it outside the organization.

This is exactly what Volker wants to do. Right now, he is an individual with his own private modified version of a GPL program. He's looking for a team, and he's saying that he is only interested in team members who agree not to release the program. What's the problem?

Tord

[Gian-Carlo Pascutto]

This is exactly what Volker wants to do. Right now, he is an individual with his own private modified version of a GPL program. He's looking for a team, and he's saying that he is only interested in team members who agree not to release the program. What's the problem?

Tord

Hi Tord,

I agree it is legal. See for example this very similar case:

http://lwn.net/Articles/178550/

Now, whether it's a good idea...we will see how soon (if not already), you will find "private Glaurungs betas" being sent to testers without source. Good luck explaining why that is illegal and this is not

[GenoM]

hi Tord, my respect, but announcing publicly "private GPL betas" and "looking for "trusted" testers", is imo nonsense.
Who are these mythical "trusted" testers? And who are "untrusted" ones in this case?
In any case, i'm not going to apply for such testing

[swami]

Glaurung 2.2 NNT is another glaurung clone (pardon, derived work).

Changes
- Simplification (Bugfix?) in LMR
- Bugfix in move ordering

Expect it to be 5 or 10 ELO stronger than Glaurung 2.2.

If you want to get a copy, please send an e-mail to glaurungclone [ät] nnuss DOT de.

I will send copies to trusted testers only.
If you don't get a copy, no reason will be given. This does not necessary mean, that I don't trust you.

Greetings
Volker

Am I the only one who thinks that original poster was intending to be ironic or funny?

I'm assuming that he doesn't have the derivative, atleast not prior to creating this thread.

He gave away too many clues as to how this was not serious:

Few indicators:

* the claim that it's 5-10 elo more than Glaurung. Who knows if it turns out to be -5 elo?
* Asking for interested and trusted testers (oh yeah, for an alleged +5 elo derivative?)
* Changes: not too convincing, atleast for a non-programmer anyway.
* Email address - that begins with Glaurungclone@ is a satire for someone who is looking for serious trusted testers.
* the claim that if he doesn't send the derivative to someone, it doesn't necessarily mean he doesn't trust them - implies, that he doesn't have anything to offer in the first place.
* His participation in other thread where he criticized someone who is pro-derivative.

Kudos to you for making it appear genuine, Volker. Going through this thread, it seems as if everybody assumed that this was genuine offer.

[diep]

This is exactly what Volker wants to do. Right now, he is an individual with his own private modified version of a GPL program. He's looking for a team, and he's saying that he is only interested in team members who agree not to release the program. What's the problem?

Tord

Hi Tord,

I agree it is legal. See for example this very similar case:

http://lwn.net/Articles/178550/

Now, whether it's a good idea...we will see how soon (if not already), you will find "private Glaurungs betas" being sent to testers without source. Good luck explaining why that is illegal and this is not

You skip the most difficult part of the announcement:

finding TRUSTED testers.

If you know one, forward me e-mail adress and name of 'em also.

Vincent

[Gian-Carlo Pascutto]

I'm assuming that he doesn't have the derivative, atleast not prior to creating this thread.

* Changes: not too convincing, atleast for a non-programmer anyway.

I think it is genuine. Volker showed his NN based timing in his own program last time I was in Leiden, and it looked like a very cute idea to me.