How would you feel if SourceHut required all public projects to have either a (1) FSF approved free software license, (2) an OSI approved open source license, or (3) a Creative Commons license?
@sir I'm assuming you mean the hosted version verses the code they can self host on their own resources.
But, assuming yes, I'm in favor of it because organizations who want to have open repos with other licenses can host their own services.
If no, probably still in favor but just a little less strongly. :)
Regardless, I'd like *all* public repos to have *a* license regardless of the results. I dislike limbo projects.
@sir SourceHut the instance, or SourceHut the software?
@avalos the instance
@sir Do it. It'd be great if it also had a no-CoC policy too, heh.
@sir I do think that license limitations are good in theory, but in practice I'd prefer to leave the door open to things like the BSL and the License Zero Parity/Prosperity license - things that are in the spirit, but do not conform to those definitions to the letter.
@AlexMax strong disagree, they are not in the spirit of these licenses
Excluding these licenses would kind of be the point
@sir I guess I don't understand why they're deserving of a targeted exclusion.
@AlexMax free & open source software has NEVER been about non-commercialization.
@sir BSL basically automatically turns the code into a GPLv2+ compatible license after a set date (otherwise it can't legally be called BSL). I dunno, to me that seems completely reasonable, assuming the conditions before the change date aren't too unreasonable.
I can understand being against non-commercial licenses with no expiration date, so that's Prosperity out, but Parity is a copyleft license.
@AlexMax license your shit like that if you want, but it wouldn't be welcome on my platform under these terms. These are not free or open source licenses, nor are they in the spirit of free and open source, and they are unacceptable by my standards - and by the well-documented standards of the FSF and OSI.
@sir Yeah I think this is the fundamental disagreement between you and me. I don't think the FSF and OSI should be the only arbiter of acceptable licensees in the open source community.
@AlexMax and I don't think you're a member of the open source community at all. You're a member of some other community, because what you're advocating clearly and objectively is not open source, and you're mounting an active attack against something I hold dear.
@sir I think you're confusing me for somebody else. All of my open source projects thus far have been released under OSI-compatible licenses, and the open source projects I've contributed to have - to the best of my recollection - also been under OSI-compatible licenses.
What about licenses that have partial approval by OSI such as Unlicense. It is considered free but not recommended by OSI. I have worked on projects that use this license where would they fall?
@chmod777 aye, if it qualifies for the OSD it would be permitted. Does not need to be recommended
@sir How would this be enforced? Making you pick a license from a list? Having an automated process that checks for a license file that matches an approved one?
I'm mostly asking because I like to dedicate projects to the public domain, usually by putting a dedication in the README. It doesn't legally count as a license since I can't provide a license for a work I no longer hold a copyright for, and it's not super machine readable, but it still gives people all the rights needed to be FSF and OSI approved.
@montagsoup I would put it in the terms of service and then reach out to projects who are out of compliance as it comes up.
Use CC-0 if you want to public domain your stuff
@sir I don't think most users would agree, but if you made something like "open source portal" that required all of its projects to use OSI approved licenses, then yes.
also, would be interesting if you shared the most popular licenses used by projects in sr.ht
@sir The FSF and the OSI have good intentions, but I'm extremely wary of organizations taking it upon themselves to assert what is or is not a "free" or "open-source" work. In reality, people choose a variety of ways to make their software free, and peoples' choices shouldn't be restricted by arbitrary authorities.
You might agree that copyright law has gotten out of hand in recent years, but the FSF & OSI take advantage of copyright law just the same to assert their interpretation of FLOSS.
@swashberry the OSD and four freedoms are both good models, and mostly equivalent, having two organizations maintaining separate (but largely identical) lists of licenses helps to keep the system honest.
The compliant here seems kind of philisophical and not rooted in fact
@sir That's an arguable point, but you didn't use the OSD or Four Freedoms, you used the word of the FSF and OSD. Just to provide one example, Coraline Ada Ehmke, a person who seriously and explicitly advocates for an anti-meritocracy stance, was a serious contender for the board of directors for the OSI.
In the end she didn't end up getting in, but if she *had*, and you had named the OSI as an authority figure on what is or is not open-source, you would have seriously fucked over your users.
@swashberry Coraline's licenses do not meet the OSD or four freedoms, and it's incredibly obvious by anyone's reading. If she had been elected she would have had to *change* the OSD.
If the OSI or FSF stops being a good steward of these definitions, then we can change the policies on SourceHut. For now, I think they're doing a fine job.
@sir I have serious issues with the Free Software Foundation as well, and their protectionist demand that people cough up their source code. The free market of ideas doesn't demand that people reveal *how* they got a solution, merely that a solution exists. By requiring that people reveal their source code, the FSF demands that people surrender market competition in favor of homogeneous compliance to, again, a completely arbitrary authority figure, which to me seems very not-free.
@swashberry I reject your axioms, access to source code is important and the entire point of the idea presented here
@swashberry oh, and I don't believe everything in life is driven by market forces
@sir Not by market forces, no. We're all individuals, after all. But what choices we have, what options are available to us, is subject to the will of those around us, at least as long as we live in an authoritarian society. Market forces are those created as an emergent property of those around us: what products are available, for what reasons, et cetera, all as a result of social (or legal) pressures from our peers.
@sir Ultimately what we're trying to do is find the use-case which benefits the most people, and it seems silly to me to say that isn't the one which benefits the most *individuals*, since it's individuals and their influence which shape the market as a collective.
@sir Access to source code is already implied by the fact that the source code is hosted in a public repository.
The question I'm posing, and you can criticize this for being philosophical if you like, is what made you found SourceHut in the first place? Apart from a different implementation, what was your motive? Was it to create a platform where your works wouldn't be restricted arbitrarily? If so, would it be better to arbitrate what is FOSS yourself, or to grant more freedoms to your users?
@swashberry @sir unlicensed or non-free licenses on code doesn't make it available even if it's in a public repo. In those cases you can't fork it, you can't include it in your projects, you can't even really *read* it because you might be the target of patent trolls if you write similar code later.
You need *a* license to begin with, and it needs to be free in order to freely use the code in any way (including just reading it).
In the good old days, copyright was an opt-in system where you had to demonstrate that your work was original and you wanted it protected. You didn't have to go to insane lengths to specify that you want people to be able to use information which is freely available. Unfortunately those are not the times we are living in now
@swashberry @sir my problem with this idea is that modern capitalism is mostly based on competition as the only means of improvement.
When humans as a civilization have used cooperation as the means of our improvement throughout our history.
In a certain way, free software licences are a stance for cooperation instead of competition, and I think it works.
@lorabe @sir I don't really know where you got that idea. Cooperation happens a lot in capitalist societies; they just also include competition, and that's important because competition drives innovation.
Competitors challenge one another to find the best solution for a given use-case and cooperation helps to achieve that use-case.
It's a balance thing, and I think the balance is disrupted when cooperation is enforced and competition is hindered.
@swashberry @sir cooperaron is naturally limited by patents and intelectual properties in a market based on that, and while competition improves things, it rarely creates it, so innovation is not necessarily driven by competition.
Society improves when they acknowledge their needs and then they can cooperate or compete to achieve that.
Free software licences functions well because the cooperation is not limited by the competition, you can compete as much as you want but not to punish others.
@sir I feel divided. I like the idea in theory, but I'm just not sure if the lists/definitions are "good enough".
I feel like it is certainly reasonable, but I'm not sure if it would result in better or worse practices.
@eldaking what sorts of criteria or licenses are omitted by the OSD or four freedoms
@sir It's not a specific thing.
It's more like "people might be interested in using a really cool license that isn't recognized by the foundations, but instead they choose one of the famous but not ideal licenses".
Particularly as the situation change over time (I'm thinking things like the GPLv2 not having provisions for compatibility with later versions, for example).
@sir I can see the benefits. People won't screw up by using licenses that are worse. But they also can't (theoretically) create or adopt better licenses.
(Not because the FSF/OSI/CC are particularly bad, but because that's the downside of restricting choice to a subset).
And I, personally, don't feel like I know what is more important in this situation.
@sir This could have an educational benefit. Many people do not understand that open source software requires more than simply publishing their source code.
I frequently come across small projects on Github that have no license anywhere in the repository. Often they're an interesting foundation I might like to fork and build upon, but if the original author is unreachable, there's nothing to be done.
GitHub provides a few gentle nudges, but making it a hard requirement might be a better move.
@sir Contrary to some of the other replies in this thread, I think your proposal may not go far enough. Being a bit more choosy about what licenses are acceptable could yield further dividends.
One benefit of open source software is code reuse. Incompatible licenses threaten this. I think Creative Commons made a huge mistake creating the NC licenses, which can't be combined with e.g. BY-SA.
I would consider picking a subset of licenses so that all code on the platform is license-compatible.
@sir it would be easier to vote if there was a list of licenses approved by any of these bodies
Imagine you want to mirror some third party project under a strange license.
@sir "free" means "I can pick any license for my projects". If this is not the case for sr.ht, then I'm not going to use this instance.
Also, "FSF or OSI approved license" means empowering these organizations. I despise them.
@sir my understanding is that by default, ie before you choose a license, it is copyrighted.
that would be "source-available"? ie, you can't publish modifications, correct?
Remember when you sent your GNU ethical repository criteria self-evaluation? You argued against forcing users to use free licenses (A4), and you said:
> I also reckon that source-available software is better than
proprietary software, so de-platforming source-available software would
just increase the amount of proprietary software out there.
Has your opinion on this changed?
@sir It's actually a pretty big legal trap to have a public project up on a code hosting site, with an open-source aesthetic, but with a license which doesn't provide the freedoms we expect from an open-source project. One great example is this fucking project: https://github.com/AU-COVIDSafe/mobile-android/blob/master/LICENSE.md
Leaving projects like that up, or projects without a license (hence "all rights reserved"), is just asking for people to accidentally break the law.
@sir As to whether to rely on the OSI's OSD and the FSF's recommendations... This one seems pretty simple to me, those orgs have been reliable in defending the term "open source" from corporate interests, so relying on them makes sense.
If that triggers people too much, you could have a way to suggest a license, where people make a case for why it should be valid for a sourcehut project. Have a page dedicated to explaining why already-shot-down licenses aren't available. That's more work tho.
@sir Would it also apply for unlisted repositories?
@sir There are loads of hosting options for projects that don't fall under an approved OSI/FSF/CC licence. A forge with a policy of open sharing of code and content under terms that guarantee freedoms and filters out questionable licences? Yes please.
@af no, SPDX aims to list comprehensively with no particular criteria or defense of any specific rights; and Blue Oak is a bad license
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!