Paying for a license would be nearly impossible, as the OBS team would need every contributor to sign a CLA to give the OBS team the rights to relicense/dual-license the OBS code base.
As long as they all pay a reasonable license fee, I don’t see a disaster. The OBS project could then pay developers to build open source features that benefit all. A lot of the forks would likely contain features that are not if interest to other users anyways.
At this point, it's not even clear to me that paying would rectify the legal issue with respect to the GPL. It's an amalgamation of source contributions over time. Any one contributor could, in theory, refuse any consideration other than an open sourcing of the software.
Of course, now I think about it, that could be an easy problem to fix. They say every man has a number.
Almost all relicensing efforts are actually most hindered by not being able to contact people. If you can't contact someone who holds the copyright to something, you can't change the licensing rights over it.
A lot of projects have copyright assignment, to allow for relicensing. They typically ask for you to assign copyright to them or to a company they control, so that they can still relicense as they see fit, in future.
It's worth noting that if someone can't be contacted, the maintainers aren't out of luck yet. If the contribution is deleted (and then possibly reimplemented later by someone with whom the project _is_ in contact) then the issue is resolved. It can be a lot of work, though, depending on the size and importance of the contribution, and reimplementing the code in a way that doesn't derive from the original submission can be difficult or ambiguous.
I guarantee you every big tech company "steals" GPL code without intending to steal. Most end up amending their mistake when they realize they misused a license.
You can legally dual license if you have ownership over all the code in question. This is common for open source industrial software; pay for a different license so you can embed it in a closed source project. The GPL doesn’t restrict you from offering the same code with a different license if you own it.
Often the issue is that some projects don’t require contributors to sign over copyright ownership as part of contributing. So you have a project that’s licensed uniformly, but each contributor still owns their individual contribution. Unwinding this after the fact can be a nightmare, as it involves either finding every contributor and asking them to sign over their code, or manually removing every bit of code you don’t own as a project.
This is why a lot of bigger projects require you sign a contributors agreement that assigns copyright before you can contribute to the main repo. Doing this in advance saves the project a lot of headaches down the road if dual licensing is deemed useful. This is true even if you want to license under two different open source licenses, as only the copyright holder can change the license.
The GPL itself offers no such option, but if the copyright holder(s) choose to they may offer whatever alternatives they choose. Many significant open source applications are offered under this model such as MySQL.
The catch is that the more copyright holders there are the more likely it is that someone who has contributed a non-trivial part of the project will not agree, in which case their work would have to be removed/replaced to allow for relicensing.
Large projects that have not required a CLA from contributors are effectively impossible to relicense.
Basically, it's practically (not ethically) fair game to use GPL in commercial software, until someone catches you. The only repercussion that the license provides is that your license is revoked until you resolve the violation (for the first violation).
Meanwhile, you got to release your product, and by the time you got caught you've had enough time to implement it yourself.
The authors of the software can create an additional licence that they could buy. Of course, that gets more difficult if there are many different, hard-to-contact authors.
I don't believe you need to demonstrate specific damages for a copyright infringement case in the US. You only need to demonstrate two facts: that you are the legitimate holder of the copyright, and that the other party did in fact infringe.
It's pretty funny that corporations can levy a multi-million dollar judgement against a single mother for pirating a CD, but then when the tables are turned, it's no big deal
As others have said, penalties are not related to how much you charge for the product itself. For example, when you start illegally distributing music your penalty won't be retail cost * number of copies. There's multipliers & things that get applied. Basically your judge/jury will figure out the damages amount after you're found guilty (assuming you don't settle).
This also makes sense when you factor in that retaining lawyer services to prosecute the infringement costs time and money (not to mention the court's time & resources to handle the case).
Suppose, I'm an author of GPL software. I think that my code costs $1M. I expected that if someone uses my software, according to license, then he will release his software under same license for me. Now, somebody used my $1M project in his $100M project in violation of my GPL license. My losses are $100M.
Only if the project owns the copyright or otherwise has been granted such powers in their contribution agreement. Otherwise, no. They’d have to get approval from every autho/rewrite the code they don’t have a license for to provide a copy that isn’t GPL.
The GPL contract only says you have to distribute source to the users you give binaries to. The only people who can ask for said source are the people receiving those binaries and the only people who have standing to sue when that doesn’t happen is the copyright owners.
That’s why you can use GPL software in your private CI system and not need to give anyone the source code.
