Comment by TheDong | Hacker News Reader

Comment by TheDong | Hacker News Reader

Parent Full thread

I'm not a lawyer, but looks totally fine to me.

If you look at the link they have for proof, the change was GPLv3 to a dual-license AGPLv3 + not-really-specified license you can privately arrange.

They have to respect the original GPLv3 license, which means that Core has to continue to publish all libpebble3 changes under a GPLv3 compatible license, and they do appear to be doing so, even if they also offer a separate license for sale.

I feel like rebble is phrasing this a little misleadingly too. The neutral phrasing here would be "Pebble forked our work, and per our GPL license is continuing to make all their changes available to all users for free. If you contribute to their repo, not ours, they now require a CLA, and for code they write you can also pay them for a difference license (though it's always also available for free under the GPL)"

There may be something that's real here, but "forked our library and added a CLA" feels normal and expected, not worth hostile phrasing.

Wait a sec. IANAL, but if I license something to you under the GPLv3, you may not license it to someone else under AGPLv3 or a commercial license.

That being said, libpebblecommon seems to be Apache 2.0. But this part of the diff seems questionable:

> # Copyright and Licensing

Copyright 2025 Core Devices LLC

How does Core own the copyright to this code?

The AGPLv3 is for the new code core writes going forward I would assume.

Distributing a mix of AGPL and GPLv3 code is pretty reasonable to do, right, and I think basically all the user's rights under the GPLv3 are being fulfilled just fine.

I agree the commercial license could be dicey, but I assume in reality it's the usual AGPL thing where it's "If you pay, you don't have to comply with the network-services bit, but you now get the code under the GPLv3, so you have to make a network service and ensure your users _never_ get binaries containing this code".

Or, possibly even more realistically, they've put that there and if anyone says "We'll pay $3M for un-encumbered code" they'll rewrite the code from scratch to make it un-encumbered by the old GPL code, and until someone says a number big enough to cover the rewrite they'll never actually do anything.

> Copyright and Licensing

A forward-looking section applying to all new changes going forward I guess.

As long as they've preserved the old copyright notice somewhere, and it's given to users who request it, it doesn't really matter what the README says does it?

I promise I'm not a shill for them. I do think what they're doing comes off as overall not great, but not as "willful GPL violation" (they're still sharing code), and not as egregiously malicious as the blog makes it sound, so the blog author has me a little unsympathetic with their own misleading (in my opinion) phrasing of this stuff.

> Wait a sec. IANAL, but if I license something to you under the GPLv3, you may not license it to someone else under AGPLv3 or a commercial license.

Not exactly the above case, but from the GNU GPL version 3 (https://www.gnu.org/licenses/gpl-3.0.txt):

      13. Use with the GNU Affero General Public License.
    
      Notwithstanding any other provision of this License, you have
    permission to link or combine any covered work with a work licensed
    under version 3 of the GNU Affero General Public License into a single
    combined work, and to convey the resulting work.  The terms of this
    License will continue to apply to the part which is the covered work,
    but the special requirements of the GNU Affero General Public License,
    section 13, concerning interaction through a network will apply to the
    combination as such.

Since it's a more restrictive license, they can't merge back the changes.

Can't they? They're given the option to take the upstream code under the AGPLv3, so if they take the code as AGPLv3, they can incorporate their changes since AGPL and GPL are compatible.

Well, yeah. I'd never really thought about it before, but linus was really prescient in not adopting the or-later clause. This sort of thing would be destructive to linux.

Reply on news.ycombinator.com