✧✦Catherine✦✧
whitequark@mastodon.social
another case of AGPLv3 being used extractively! https://github.com/minio/minio/commit/27742d469462e1561c776f88ca7a1f26816d69e2
2
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@joshix legally speaking, as best as a i can tell not being a lawyer, no.
0
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @joshix no, but they compel you to license your contributions under Apache-2 to "MinIO maintainers" in order to get your PR merged; basically a CLA-without-a-CLA
2
1
0
@whitequark This era of "AI" is such an unmasking event for people who would totally sell you out for that sweet VC money.
0
0
0
@whitequark @shironeko @joshix So it’s effectively a plain old permissive license.
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@airtower @joshix @shironeko no, the Apache-2 grant is _only_ to "MinIO maintainers", not to anybody else
0
0
0
🆘Bill Cole 🇺🇦
grumpybozo@toad.social@whitequark @shironeko @joshix FWIW, ASF also requires CLAs for all contributed code (obviously also under Apache-2) purely as an ass-covering mechanism so that the rights to publish/reuse/modify the code are clear. It’s not clear to me that any open source project can safely accept contributions without some sort of CLA, even if it is purely implied.
1
0
0
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @joshix @grumpybozo Linux has a Signed-off-by: mechanism, which I think Bill refers to as a type of CLA?
1
0
0
@whitequark @shironeko @grumpybozo CLAs generally have several parts, Signed-off-by replaces only one (you representing that you have the right to contribute that code), which is the least controversial one but also arguably the most important one after the whole SCO copyright kerfuffle.
0
0
0
@whitequark I don't understand how this relates to the AGPL - any maintainer of any project might do something like that, the AGPL like various other licenses allows you to fork and continue. What's the specific license failure mode?
1
0
0
@gabe the primary function of AGPL is to enable commercial companies to extract fees from their users before pulling shit like this. that's what it's used for in the real world. there are some other AGPL applications (like Mastodon) but they're a rouding error in comparison in my assessment
2
0
0
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @gabe the AGPLness does matter: the fact that it is OSI-approved means you can milk the community for contributions by pretending to be FOSS until you rugpull
1
0
0
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @gabe I don't see how this hypothetical is relevant?
1
0
0
@whitequark I don't understand. I'll try to figure out more myself (explanation from you would be welcome, but I don't want to impose).
My understanding is that the AGPL adds nothing at all to the GPL that makes this behaviour easier to pull off or different in effect or harder to work around. Apparently, that understanding is incorrect, because you're pointing specifically at the license, not just the shitty actors being dicks. Clearly, I need to look harder.
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@gabe the AGPL is incredibly broad and by virtue of this makes compliance by people running stuff in the cloud very difficult. if you run a GPL'd service X in the cloud, you have no obligation to do anything, so no incentive to care. if you run an AGPL'd service X in the cloud, you're suddenly potentially legally liable, which incentivizes you to buy a commercial license as essentially a form of insurance against legal action
2
0
0
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @gabe well, the scenario you're describing doesn't happen* and the scenario with AGPL happens over and over. I'm a lot more interested in that than in hypotheticals
* actually i can think of one example, Douglas Crockford's "Don't do evil" license. he ended up granting IBM an express license to do evil for free and then bragged about it in a talk
1
0
0
Charlotte/Cinny 
charlotte
@whitequark @gabe the choice of licensing also makes it potentially not legally feasible to ship it modified — no good way to prominently offer users copies of the source code without diminishing the value of minio as a file storage service
1
0
1
✧✦Catherine✦✧
whitequark@mastodon.social@charlotte @gabe yep, this is a big part of _why_ the incentive works
a different OSS license would not have the same effect; the AGPL is useful for extracting rent for precisely the same reasons why it is considered useful for the GNU project
0
0
1
Everything else you said completely agree.
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @gabe I subscribe to POSIWID: if the practical impact of AGPL is toxic corporate bullshit, then that is something that is wrong with AGPL regardless of what anybody's intent is
1
0
0
@whitequark okay, yes, I understand that part, the license's entire idea is to somehow expand "distribution" to also include "run as a network service", to get Faceglezonbay to contribute/share extensions or whatever, and I'm aware of the criticism around both that idea and the specific AGPL approach.
My confusion is how that relates to minio's specific "we're done here" commit you linked at the top. It doesn't seem to make any difference either way to anybody who used or uses the codebase, which was already AGPL'd. The weird contributor licensing thing is weird, but that it effectively allows minio to quietly dual-license the product (directly where they have the copyright, via the unrestricted license they get from others), it's also not functionality different from what others do with their projects.
I understand criticism of the AGPL as such, just not what it has to do with the "maintenance mode lol" commit.
1
0
0
@gabe the commit is evidence that minio's use of AGPL was in bad faith in first place: they have not used it to share work with the OSS community, but solely to extract rents. AGPL enables such extractive use by virtue of its text. once they wanted more rents, they've privatized the project.
the CLA-like contribution policy is further (and stronger) evidence towards that, but I was unaware of it when I've encountered the commit
0
0
0
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @gabe were AGPL not an existing license that is accepted by OSI and us as something that's OK to contribute to, they'd have used some sort of source-available license and we would've collectively avoided contributing or depending on that shit in first place
1
0
0
1
0
0
1
0
0
✧✦Catherine✦✧
whitequark@mastodon.social@shironeko @gabe a full 30% of those involves AGPL! I think describing this as "hardly relevant" is a bad faith tactic so I'm not really interested in continuing this conversation
0
0
0