Learning More About KDE's Plasma Next Desktop

[erendorn]

I think your lack of legal knowledge is contributing to the many errors you make in your assessments. As said before, the FSF copyright assignment is also a CLA. The point being that FSF only accepts copyright assignments from certain licenses (or at least, will re-license to a proper FSF license if they ever should get unlicensed code assigned.)

So the fact is, that any transfer of copyright to FSF is also a CLA where the contributor also sign his acceptance of the e.g. GPL 3.0 license while transferring copyrights too. It is quite simple really, FSF only accepts code where the contributor have signed a license agreement (CLA). GNU.org have lots of information on this in case you are genuinely interested.

As said before, this CLA gives FSF projects legal protection against hostile legal attacks and ensures the code will always remain Free, just like the Qt CLA. GTK+ and many other projects doesn't have that protection.

Is it actually possible to unlicense some code? I remember someone tried once, but did not went through with it in the end.
In any case, if the license was to be kept valid in any single jurisdiction in the world, couldn't a third party redistribute legally from there to any other jurisdiction?

[interested]

Is it actually possible to unlicense some code? I remember someone tried once, but did not went through with it in the end.
In any case, if the license was to be kept valid in any single jurisdiction in the world, couldn't a third party redistribute legally from there to any other jurisdiction?

A complex question, but the short answer is "no", but the longer answer is "maybe". If an original author have released the source code to a program called "foobar v1" with e.g. a GPL 2.0 license, he has in effect made an CLA, signing away certain IP rights according to the wording of the license and local law. He can release another, identical version of "foobar v1" without the license, but he can no longer prevent the distribution and use of the original program under the GPL 2.0 license. This is a major force with the GPL license, that it allow the source code to remain Free no matter what.

There are some ways that the original author of "foobar v1" could legally remove the GPL license however. They would usually be the same methods that allows people escape from other normal signed contracts. This is not easy, since contract law around the world generally say, that a contract is a contract and must be upheld, even if it very disadvantageous to one of the signers.
One such loophole however could be if the original author in fact didn't have the right to license the program. Some companies have contract clauses that stipulate that all such creative work (IP) made by the employee, even if made in his own spare time on his own equipment, in fact belongs to the company. Such a company could in fact claim IP ownership of "foobar v1" and nullify the original authors license. It could then be a copyright infringement to distribute the "foobar v1" program, even if it had a GPL license attached.

This is why some projects like the FSF demands CLA's where the signer explicitly says he owns the program and has agreed to the FSF license and transfers copyright to the FSF, since such CLA's provide a really good legal shield against many of such loopholes, and provides good legal protection for the contributors against legal attacks.

Regarding distribution, then again the answer all depends. Most countries have signed the Berne convention about IP rights, but even a loophole in one country, doesn't make it legal to distribute to other countries where such loopholes doesn't exist.

[erendorn]

A complex question, but the short answer is "no", but the longer answer is "maybe". If an original author have released the source code to a program called "foobar v1" with e.g. a GPL 2.0 license, he has in effect made an CLA, signing away certain IP rights according to the wording of the license and local law. He can release another, identical version of "foobar v1" without the license, but he can no longer prevent the distribution and use of the original program under the GPL 2.0 license. This is a major force with the GPL license, that it allow the source code to remain Free no matter what.

There are some ways that the original author of "foobar v1" could legally remove the GPL license however. They would usually be the same methods that allows people escape from other normal signed contracts. This is not easy, since contract law around the world generally say, that a contract is a contract and must be upheld, even if it very disadvantageous to one of the signers.
One such loophole however could be if the original author in fact didn't have the right to license the program. Some companies have contract clauses that stipulate that all such creative work (IP) made by the employee, even if made in his own spare time on his own equipment, in fact belongs to the company. Such a company could in fact claim IP ownership of "foobar v1" and nullify the original authors license. It could then be a copyright infringement to distribute the "foobar v1" program, even if it had a GPL license attached.

This is why some projects like the FSF demands CLA's where the signer explicitly says he owns the program and has agreed to the FSF license and transfers copyright to the FSF, since such CLA's provide a really good legal shield against many of such loopholes, and provides good legal protection for the contributors against legal attacks.

Regarding distribution, then again the answer all depends. Most countries have signed the Berne convention about IP rights, but even a loophole in one country, doesn't make it legal to distribute to other countries where such loopholes doesn't exist.

Thanks for the detailed answer. Does it differ among open source licenses (in particular, would it be the same conclusion for shorter/less detailed licenses such as the permissive ones)?

I don't think I agree on you CLA's point though. If the signer says he owns the program but actually did not, his signature is worth nothing. Wouldn't the CLA would be as void as the license?

[interested]

Thanks for the detailed answer. Does it differ among open source licenses (in particular, would it be the same conclusion for shorter/less detailed licenses such as the permissive ones)?

This is all down to particularities of the individual license, but the general answer is, that even very permissive licenses or less detailed licenses can't be retracted as such even by the IP owner. But badly formulated licenses may leave lots of legal loopholes, so it is a very good idea to use eg. unmodified FSF licenses. (IANAL etc.)

I don't think I agree on you CLA's point though. If the signer says he owns the program but actually did not, his signature is worth nothing. Wouldn't the CLA would be as void as the license?

Yes, you are in principle right in the scenario you describe; one cannot release a program under any license if you don't own the IP.
But what you describe is a de facto rule that the CLA signer did in fact not own the program. This is exactly what would be contested in a lawsuit.

A signed FSF CLA is still a powerful legal weapon in such a dispute. This is complicated, but lets take the previous example, where a company try to claim IP ownership of program made by an (former) employee in his own spare time.

Without the FSF CLA the company's lawyer would have it easy; just the threat of a long and expensive lawsuit may make people cave in an sign a eg. IP transfer right uncontested. Or they will face a (cheap) local lawyer perhaps without much IP experience. But if you are up against the FSF whose lawyers are all expert in software IP, they will face tough opposition on every claim.

It is important to remember here, that whether the company actually have the right to such IP isn't a given thing, but is only a claim. Maybe their original clause stated that they owned all such IP, but is that clause actually legal in the local jurisdiction?
With a signed CLA you have strong bulwark against unmerited claims, and you ensure that no case is an easy pushover.

Furthermore, you can also limit any damages even if in fact the claim against program IP had any merit. A signed FSF CLA is a strong legal evidence of good faith, and just the fact that there is a legal entity to negotiate any damages (monetary or not) in court are likely to reduce the scope of such damages, like preventing the claimant from issuing DMCA take down notices.

To sum up; a signed FSF CLA is strong protection against unmerited claims, ensures that no legal bullying tactics will win, and that any claim will meet stiff opposition against expert lawyers. This alone will prevent unmerited legal attacks to even begin. But it will also reduce the scope of damages, even if the claim in fact had any merit.

[erendorn]

But if you are up against the FSF whose lawyers are all expert in software IP, they will face tough opposition on every claim.
This alone will prevent unmerited legal attacks to even begin. But it will also reduce the scope of damages, even if the claim in fact had any merit.

Ok, it will indeed limit legal bullying.

[lolnope]

This is a taskbar mockup made by the kde design team.
They are now working on a new cashew icon
Things are looking good I guess

[lolnope]

This is a taskbar mockup made by the kde design team.
They are now working on a new cashew icon
Things are looking good I guess

Eh.. For some reason the image isn't displaying for me on Firefox android
Anyways here's the link

http://wstaw.org/m/2014/03/03/taskbar_1.png

Also stop arguing about the cla. It isn't going to change anything if either of the side wins the discussion

[Luke_Wolf]

Eh.. For some reason the image isn't displaying for me on Firefox android
Anyways here's the link

http://wstaw.org/m/2014/03/03/taskbar_1.png

Also stop arguing about the cla. It isn't going to change anything if either of the side wins the discussion

more like stop responding to Honton there's no point, and it's best to just ignore the fantroll.