Shoutout to all of the contributors to SourceGraph whose work was just made closed source, big thanks to SourceGraph for pissing on them, I'm sure they really appreciate it. All those users who thought your FOSS product was really cool surely still think your product is super cool now that you've taken their rights away
Big reason I don't like open core models is not just because they cause nonfree software to proliferate but also because it's a pretty clear promise, especially when combined with a CLA, that they will pull the rug at the first sign of a bad quarter.
@devnull you teach those lawyers how the GPL works? Bonus guess: you sell the relicensed version to corporations and contributors from the community don't get a portion of that sale?
@drewdevault I unfortunately don't get to talk to the corporate lawyers. It's either I lose the contract or implement the CLA. If I asked I'd probably be billed $600 an hour just for them to say no.
Re: sale of relicense — yes. I'd sooner rather not do it at all, though.
If #MIT works better a re-license may be in order. I don't know how corpos feel about MIT, but I imagine it's ok...
@devnull I don't have any problems with a company publishing free software or selling dual-licensed software, until they start accepting the free labor of their communities and selling it as if it were their own, and retaining the exclusive right to commercialization of the product. If you aren't going to extend your community the same rights they extend you, then do all the damn work yourself.
FYI selling GPL'd software works fine, see for example, uh, Linux
@drewdevault I'm not sure why you're being so aggressive right now.
There's no exclusive right to commercialization with #NodeBB, none is expected, and someone could start a competitor with our software if they so wished.
I reached out to try to have a civil discussion about our rationale for requiring a CLA, and to explore alternative solutions, but you seem more interested in shaming others instead of talking.
@devnull sorry, I didn't mean to be impolite. The essence of my comments is true, though, so long as you aren't the sole author I don't think it is reasonable to assert the right to sole commercialization. Hope that clears it up!
@devnull@drewdevault I'm not sure I understand the question. You are asking that if the product were licensed under Apache, MIT, or BSD, would you still get questions from companies thinking about using it about the license? Or something else?
@soaproot@drewdevault yeah that's what I'm asking. I think the viral clauses of GPL make it a non-starter for some corporations, but a license like those above would be easier to pass through legal.
People call the GPL "viral" as if copyright isn't already viral. Use one work in another, and the original rightsholder now has a claim to the new work. That's how copyright works. The GPL doesn't change that. You'd still need to talk to the original rightsholder if you wanted to redistribute anything.
The only thing GPL changes about this is the addition of "if you don't contact the rightsholder to negotiate, here's something we already agree to"
@wpalmer@devnull@soaproot@drewdevault The comparison is with other free software licenses. Free software was already a thing before RMS came along. A lot of it was in the public domain.
@devnull@drewdevault Sounds about right. I haven't paid super close attention to the policy at $DAYJOB but I think it is some version of "if it is Apache, MIT, or BSD you are fine otherwise fill out this form and blah blah blah" (probably an oversimplification, but something sorta like that)
It's especially egregious with permissively licensed stuff. Like the only obligations the maintainers have towards the contributors in the absence of a CLA is giving the contributors credit for their fucking work and that's too much for them!
@drewdevault I don't do CLAs unless I'm being paid for it. Such nonsense. And I try to practice the flip side too. That's a large part of what went into making #musl MIT. I did not want a dynamic where I had asymmetric authority to relicense other ppl's contributions but they not have same for mine.
@dalias@drewdevault I would have no problem with a CLA that only allows the original authors to change license from LGPL to GPL to AGPL, or update to newer versions of the *GPLs.
But isn't LGPL -> GPL -> AGPL already basically allowed by the licenses (by mixing it with new code under the more restrictive license)?
And what if the original authors get bored of the project, it gets picked up by new people, and now they don't have the right to update it to new GPL versions, which they would if that clause had been given to everybody rather than just through a CLA.
@valhalla@drewdevault@dalias No, it would provide additional flexibility for the original author - for example deciding if and when to switch the entire codebase from GPLv2-only to GPLv3-only and perhaps v4 in future.
Edit: No, sorry, not if it's "-only" rather than "or later". But this is exactly the asymmetry of power I was talking about not wanting and refusing tp create. So I'm confused why you're bringing it up as if it were an exception.
@dalias@federico3@valhalla@drewdevault Reminds me that you have things like KDE's licensing policy where they have or-later but with the requirement of being approved KDE e.V itself.
I take it as "(L)GPL-*-or-later" with overruling the FSF and keeping the relicensing rights to KDE e.V itself, I don't like it but compared to a copyright assignment it's at least public licensing and KDE e.V there doesn't ends up being a full on rights holder.
It's one of the reasons why I tend to prefer weak-copyleft/permissive licences for large codebases so a gradual relicensing can be done with no overruling of author's rights.
Add comment