So, no, it is _not_ time for a “CopyLeft AI Data License.”
First, tell me about your intellectual property philosophy. Tell me about your social and societal goals. *Then* we can talk about if #copyleft is a suitable tool to help advance your agenda.
@peepstein@xek yes, this is a very valid ideology that many advocates of BSD-style licenses have adopted (e.g., ⬇️).
The reasons that communities (like those who build FreeBSD) adopt policies that exclude, in various degrees, copyleft licensed software are principled ones—even when the outcome is software that does not respect the freedom of end users (e.g., macOS).
@peepstein@xek grsec is a good example of how copyleft doesn't guarantee any sort of meaningful cooperation between downstream derivatives and the originating authors in getting patches merged.
@xek the dynamics that make #FOSS a powerful force will continue to provide natural incentives for cloud providers (like Linux distros) to invest in upstream development.
Those dynamics are most powerful when there is an independent, self-governing community (like Linux and PostgreSQL).
When single for-profit firms require CLAs and give no meaningful opportunities for others to participate in governance, the forces are far weaker...
@xek MIT, BSD, and Apache licensed software is all Free Software. It just isn't under a Free Software license that protects Software Freedom for downstream recipients of the software, or derivatives of the software.
All that said, I am (in my personal time) a #copyleft advocate because I want to advance and protect Software Freedom for all software users.
@sjvn@webmink@dirkhh@osi@lightweight@larsmb I think RMS does not see the "open source" banner as "anti-copyleft", but rather that it focuses on the wrong thing (commerce and economic efficiency, rather than freedom).
That said, one might not come to the conclusion that copyleft is a force for good if one does not see the objective to be "providing and protecting Software Freedom for users."
That said, someone wise once wrote on the Internet:
""" if licenses are the constitution for communities like Moglen says, this act of telling some communities they don't count as open source is patronizing and colonialist. And that's OSI's job :-) """
There are licenses that go pretty far to try to neutralize software patents. In some ways, other licenses go farther than AGPLv3 does on patent rights.
@OrdinaryWonder Indeed, the obligation to allow the owner of a "user product" that contains GPLv3 licensed software (among other licenses) the instructions to build the source code, and ability to install a modified version of that software on their device, is one that many device manufacturers don't want to undertake. GPLv2 isn't materially different in that.
@OrdinaryWonder I don't think it's "hostile to big business" to offer an often attractive bargain, roughly: "If you want to use the software we make available to all to use in any way they'd like, you need to pass the freedoms forward to others that receive the software from you."
Agree that people should use whatever license they want for their work. As a user, you can decide to accept those terms, or not use the software.
@OrdinaryWonder They have the freedom to sell Linux, and software built with GCC, and many other creative and useful software works that have been freely given to the #FOSS Commons.
They don't /have/ to use those software packages.
They're free to do so, so long as they understand and comply with the obligations. I don't see how that's hostile.
I remember the #FOSS community reaction claiming that Google used their #OpenSource hosting service, code.google.com, (eventually @killedbygoogle) to bend perceptions on licenses in ways that were favorable to them.