Gaviota's EGTBs are only 6.5 Gb now

[jshriver]

Sounds exciting, anxious to see the end results.

-Josh

[Dirt]

The idea is to release the probe and generation code as open source. Most likely a BSD with no advertising as it was suggested or MIT license. I do not know the specifics of which one yet and I am not an expert, but the idea is to allow everybody to use it so testers, gui writers, and engine writers can use a common resource. For that to happen, it should be open. Whatever the people do or use in tournaments, it is a complete different issue.

BSD for the probe code sounds good, but you might want to consider GPL or LGPL for the generation code to discourage proprietary variants from popping up.

[Dann Corbit]

The idea is to release the probe and generation code as open source. Most likely a BSD with no advertising as it was suggested or MIT license. I do not know the specifics of which one yet and I am not an expert, but the idea is to allow everybody to use it so testers, gui writers, and engine writers can use a common resource. For that to happen, it should be open. Whatever the people do or use in tournaments, it is a complete different issue.

BSD for the probe code sounds good, but you might want to consider GPL or LGPL for the generation code to discourage proprietary variants from popping up.

If they use BSD code, they have to at least explain where it came from.

If they take Gaviota EGTB files and create an offshoot, what will be harmed by that? They must still obey the license. And if the license is GPL or LGPL, they can still build offshoots with a slightly different format.

I guess that only significant difference might be that they could produce an offshoot, acknowlege the source, and not publish the new code. But then who will prefer that over the original?

I do agree that the access is the most important facet if other people are going to use it (probably, only a binary is really needed for generation, but source is nice so that others can contribute improvements). And I find the proliferation of EGTB formats annoying. So as long as I can access the data from any program I would be happy. I guess that eventually, all (or almost all) programs will migrate to Miguel's format. That will put an end to the disk chewing monster in the closet.

For access, any of the following license formats would be optimal from my viewpoint:
MIT, BSD, Mozilla, ACE
because they do not put restrictions that will prevent usage in some environments.

[jshriver]

For the engine that's on your website now. Are the resulting .gtb files compressed or uncompressed?

Main reason I ask is that I'm still working on the 5men egtb's. Wondering if I should hold off till you have a final format, or if they are uncompressed if you'll release the compressor later so I can use it in the resulting .gtb files.

Thanks for all your hard work!
-Josh

[jshriver]

I'd also be willing to do any testing that you might need under Linux.

-Josh

[Dirt]

The idea is to release the probe and generation code as open source. Most likely a BSD with no advertising as it was suggested or MIT license. I do not know the specifics of which one yet and I am not an expert, but the idea is to allow everybody to use it so testers, gui writers, and engine writers can use a common resource. For that to happen, it should be open. Whatever the people do or use in tournaments, it is a complete different issue.

BSD for the probe code sounds good, but you might want to consider GPL or LGPL for the generation code to discourage proprietary variants from popping up.

If they use BSD code, they have to at least explain where it came from.

If they take Gaviota EGTB files and create an offshoot, what will be harmed by that? They must still obey the license. And if the license is GPL or LGPL, they can still build offshoots with a slightly different format.

The potential harm would be having to have multiple versions of the TBs on your computer to support different programs.

I guess that only significant difference might be that they could produce an offshoot, acknowlege the source, and not publish the new code. But then who will prefer that over the original?

Any commercial program, potentially, and if their version is 10% faster or otherwise slightly improved then so will some of their users. Importantly, I don't see licensing the generator under the LGPL as a barrier to anyone using the code, except those with an ignorant fear of open source.

I do agree that the access is the most important facet if other people are going to use it (probably, only a binary is really needed for generation, but source is nice so that others can contribute improvements). And I find the proliferation of EGTB formats annoying. So as long as I can access the data from any program I would be happy. I guess that eventually, all (or almost all) programs will migrate to Miguel's format. That will put an end to the disk chewing monster in the closet.

For access, any of the following license formats would be optimal from my viewpoint:
MIT, BSD, Mozilla, ACE
because they do not put restrictions that will prevent usage in some environments.

[Dann Corbit]

The idea is to release the probe and generation code as open source. Most likely a BSD with no advertising as it was suggested or MIT license. I do not know the specifics of which one yet and I am not an expert, but the idea is to allow everybody to use it so testers, gui writers, and engine writers can use a common resource. For that to happen, it should be open. Whatever the people do or use in tournaments, it is a complete different issue.

BSD for the probe code sounds good, but you might want to consider GPL or LGPL for the generation code to discourage proprietary variants from popping up.

If they use BSD code, they have to at least explain where it came from.

If they take Gaviota EGTB files and create an offshoot, what will be harmed by that? They must still obey the license. And if the license is GPL or LGPL, they can still build offshoots with a slightly different format.

The potential harm would be having to have multiple versions of the TBs on your computer to support different programs.

How does GPL or LGPL change this? Not in any way that I can imagine.

I guess that only significant difference might be that they could produce an offshoot, acknowlege the source, and not publish the new code. But then who will prefer that over the original?

Any commercial program, potentially, and if their version is 10% faster or otherwise slightly improved then so will some of their users. Importantly, I don't see licensing the generator under the LGPL as a barrier to anyone using the code, except those with an ignorant fear of open source.

It can also be a corporate policy that LGPL cannot be used. Ignorant fear or not, it is possible that this will prevent inclusion. I suppose that even BSD license can have the same restrictions, but I consider that even less probable.

I do agree that the access is the most important facet if other people are going to use it (probably, only a binary is really needed for generation, but source is nice so that others can contribute improvements). And I find the proliferation of EGTB formats annoying. So as long as I can access the data from any program I would be happy. I guess that eventually, all (or almost all) programs will migrate to Miguel's format. That will put an end to the disk chewing monster in the closet.

For access, any of the following license formats would be optimal from my viewpoint:
MIT, BSD, Mozilla, ACE
because they do not put restrictions that will prevent usage in some environments.

[michiguel]

For the engine that's on your website now. Are the resulting .gtb files compressed or uncompressed?

Main reason I ask is that I'm still working on the 5men egtb's. Wondering if I should hold off till you have a final format, or if they are uncompressed if you'll release the compressor later so I can use it in the resulting .gtb files.

Thanks for all your hard work!
-Josh

You can keep generating those, which will be ~38.5 Gb. The compressing program that I will release will work on those uncompressed files (.gtb) to generate the compressed .gtb.cp1 files (that is the extension name I am using now as in cp=compressed, scheme=1).

Miguel

[zamar]

Nice to see that someone is finally working for true open source EGTBs!! IMO there has been a monopoly on this thing for too long time.

[Dann Corbit]

Nice to see that someone is finally working for true open source EGTBs!! IMO there has been a monopoly on this thing for too long time.

Eugene's tablebase files are open source. It's the license that is the problem.

But Miguel's license (whatever it may be) will solve this remaining problem.

As far as the generator goes, I do not think that the license style is terribly important. But for the usage (runtime library) the license is very important if others want to use them.