I think this move by Codeberg is lowkey genius. They're trying to build a community based on a set of shared values. Those who disagree with those values are now self-selecting out of the Codeberg community, which is exactly the point of passing these kinds of resolutions.
The fact that the rule is open for interpretation means that they can effectively shut down malicious actors (i.e those who intentionally misrepresent the degree of AI assistance in writing their code) before they become a problem. It becomes an exercise in judgement instead of a potentially exploitable/loophole-able set of hard and fast rules.
There's no issue with this. This isn't the supreme court. Your income is not on the line. It's a code hosting repository. Go to GitHub if it bothers you.
Only time will tell if this set of values holds up. Personally, I think they will, and that the rules will end up being less strict than people are making them out to be, and that relatively few projects will be removed from the platform.
I’ve seen so many variations of this where a community starts adding more rules that are open to judgment.
Removing the obvious rule breakers looks like an easy win at first. Then over time the list of rules grows. Small groups become good at navigating the rule making process and getting changes pushed through. Before you know it, small groups of highly invested people are wielding the rules and the rule making process against their enemies or competitors. The community was long since cleared out from anyone who got sick of the drama, leaving a weird internal battle where different in groups are fighting with each other over little things and turning the rules against each other.
Maybe Codeberg will be different. Maybe the rule making stops here and it’s reasonably enforced against only the really bad examples, with leeway granted to others generously. Maybe it doesn’t turn into a slippery slope of little dramas that exhaust everyone who isn’t invested in the constant background drama of their code hosting site. I hope it does for their sake.
But I also think enough people like me have seen what happens to communities that go down the path of generating a lot of rules like these and will choose to steer clear of this site. It always feels like a good idea when you and all your friends are on the right side of the rules, but it’s less fun when you find yourself making choices or tip toeing around to avoid triggering the Nth round of rules that were introduced 10 steps past where you thought the rules were in a good and balanced place.
Importantly, Wikipedia's policy is a bit more detailed about what LLM contributions are and aren't allowed. There is always ambiguity, but going through specific cases and examples up front in the policy went a long way.
Not to mention what happens when these rules cease to be rules on purely technical matters. What if they make a rule against the site allowing hosting the code *of* AI agents for example?
Codeberg did make such a rule. Their announcement explicitly says that "Projects heavily tied to the LLM ecosytem" are "no longer [...] welcome on Codeberg". The code for an AI agent would definitely fall into that category.
> 7. You must not share projects that mostly consist of code written by "generative AI"-tools (including services such as Claude, OpenAI Codex). Such projects having an unclear copyright status (see requirements § 2 (1) 1 and § 2 (1) 3) and furthermore have little safeguards to ensure that they do not include harmful code (c.f. § 2 (1) 5).
I've been thinking along these lines more and more. Trying to specify exact limits and trying to account for every possible situation when making rules just encourages people to find holes and workarounds. It creates a game, a competition between the rule makers and everyone else.
When instead you state higher level values and ask for cooperation, it encourages us to work together. It's the Honor System instead of a deliberate competition between rule makers and everyone else.
Concrete example: sports
When you are playing with a group of friends and there's no referee, most everyone does their best to just play hard and follow the rules. It's not perfect, but it works for the most part because we all love the game and want to have a good time.
Once you introduce referees, the game within the game of "what can I get away with?" starts and you end up with players learning how to foul without getting caught and then players overreacting to fouls in order to make sure the rule breakers still get caught, and then it evolves into flopping to get a foul called even when there was none, and eventually people start to hate the sport altogether (see: NBA, FIFA)
Concrete example: government and police forces. Do I even need to elaborate on this one?
That's actually how EU law works! I think it's also why US companies keep getting in trouble here, they try to loophole-lawyer like at home, but it just... doesn't work that way.
Very interesting discussion, thanks for sharing that.
India has a different system where generally speaking, it behaves more like UK/US courts (common law) over civil law.
But if Supreme Court reads fundamental rights are violated then Supreme Court can read the constitution in a similar way but Supreme Court has used it sometimes to force the government hands.
For example, when the Bhopal Gas tragedy happened where people died because of the accident (Rest in peace), instead of having individual people try to sue the govt. and fight for years with no end in sight, the supreme court used article 142 to impose single lump sum settlement and centralized compensation scheme. So it can force a govt to do something.
The Sahara-SEBI Case is even more direct when supreme court forced a company to refund amount and fast-tracked it in a more civil-law style case.
But Indian Supreme court is very slow and from my understanding, does try to use it rarely and only the Indian Supreme court can invoke article 142 and its more so done to prevent genuine gaps rather than already existing provisions
I found this whole discussion really interesting
In theory this split between common law and civil law is about who you trust more: judiciary or legislature. Civil law leaves more to interpretation for judiciary whereas common law leaves very little to interpretation. There doesn't seem to be very wide consensus on which system is necessarily better to be honest.
> there's no referee, most everyone does their best to just play hard and follow the rules
Only in small groups. Any sufficiently large groups will have a number of people (possibly sociopaths) that will break the rules for personal gain. And in a lot of cases it's enough if very few people do that.
That's right, and that's when the rest of the people say "we're kicking you out", and if the offender says "but why" everyone say "because you broke the rules" and if the offender tries to rules-lawyer everyone says "it doesn't matter, we're kicking you out because we believe you are not playing by the rules".
Based on this precedence, if I'm accepted into the community today, how would I know I'll be accepted tomorrow? If you're literally just using Codeberg as a mirror, then sure, it doesn't really matter much. But like you said, that's not what they are building. They want people to invest in the community. So now, as an investor of my hard-earned free time, I'm faced with the question on whether all my investment can go up in flames on the whims of a shadow group.
You are effectively saying you do not trust the judgement of the community to accept you long term.
Why is that? Is it because you believe there would be a new policy that would exclude your workflow? What would such a policy be? Do you really, reasonably believe it would be implemented "tomorrow"? Do you think adapting to such a policy would be unreasonable?
Or is it because you don't trust the judgement of the people enforcing these new rules?
If not one of the two, what is the reason?
The point I'm trying to make here is that this just sounds like a slippery slope argument. If you don't have a concrete reason to believe you won't be welcome "tomorrow" then my mind has not been changed.
> You are effectively saying you do not trust the judgement of the community to accept you long term.
Correct. I was previously engaging with this community and had moved some repos over from GitHub. One of those repos deals with cryptography that has an intersection with "crypto" projects. Due to vagueness of the language ratified, it's unclear whether the repo now violates the ToS. This completely caught me off guard, and now my trust in the judgement of the Codeberg community has been weakened.
So it sounds like you're treating Codeberg as a neutral git host. They've made it clear they don't want to be that, they are a charity association that supports free software development by providing free software developers with services like git hosting. They are not wasting their services on people who don't align with the mission.
That seems like a completely different concern than what you originally stated? You're not worried about future policies, you're worried about the policy that was just passed.
And for what it's worth, it seems like they only care if the project is directly crypto related (e.g if you had a repo directly implementing a Monero node, that would probably not be allowed). If it's just a repo that implements a generic Cryptography concept that happens to be used both in crypto and elsewhere, then I don't see what the problem would be.
Again, the rules are judgement based. That means you can interact with the community to get clarification, and make your case that your project is worth keeping.
> That seems like a completely different concern than what you originally stated?
I started with a hypothetical. Then I used an actual real-world example. I have 100+ repos on GitHub spanning a wide array of technologies. If I moved all 100+ over, how do I know I won't fall afoul of whatever the community votes to ban next?
The point is Codeberg isn't a good fit for my situation which needs more continuity and less surprises, and that's been made readily apparent now. So I'm leaving, just like many other people who have been posting about it.
Luckily for you, as far as I can tell you don't have any concrete reason to fear having your repos taken down. Or at least I haven't been given any.
If the reason is more emotional than rational ("this feels like it's going to change more"), that's okay. Not everything needs to be governed by rigid rules. The Codeberg community knows that well. If in a few years your 100+ repos are still all fine and dandy to their rule set I'm sure they'd be happy to have you.
> Luckily for you, as far as I can tell you don't have any concrete reason to fear having your repos taken down. Or at least I haven't been given any.
Unfortunately you can't make that decision. I have to wait and find out how far the ToS violation will be implemented. Or I just avoid that stress and go somewhere else. Most people being affected seem to be choosing the latter.
> If the reason is more emotional than rational
That's a bit reductive. Risk management always has a level of uncertainty involved. No one makes decisions with 100% certainty in the outcome. Before these votes, I had a certain level of risk assessment in moving off Github To Codeberg that favored Codeberg. Now that situation has changed. It's not "emotional" to recognize that.
Why? What's so special about the place where you host your source? It's the second time I see this sentiment in the thread, and I don't get it. If an official channel tells you a project is OK, and then someone else decides it's not, what does it really happen outside of you taking your project elsewhere?
Hosting a project with its issues, packages, releases, discussions and everything is an investment. The URLs will be in multiple places, some of which I don't control.
Then suddenly the hoster decides that my project isn't wanted because the zeitgeist is against some technology used in it and poof it goes away.
What now?
"Just move away", sure. I can do that if they still let me export my data. But to be honest I'd rather just use a platform with a smaller risk of getting booted because someone decided my project is "too AI" or something similar.
I have no idea if there's a way to argue against the decisions, what proof do I need to provide, do I need to prepare somehow? Like have video of myself with a dated newspaper typing the code by hand to prove AI wasn't used? :D
>And for what it's worth, it seems like they only care if the project is directly crypto related (e.g if you had a repo directly implementing a Monero node, that would probably not be allowed).
Looking at community conversations, there actually seems to be little clarity over this, and quite a bit of confusion. Several established members of the community seem to be saying that the ToS actually only bans "cryptocurrency related projects" that "[harm] the reputation of Codeberg", and that direct cryptocurrency projects that don't harm Codeberg's reputation are actually allowed, saying that the current banner is a mistake and is in the process of being changed (while they have their own systems and are not on Codeberg, I'm guessing GNU Taler might be an example of a cryptocurrency project that would be allowed). In fact, this appears to be within the official interpretation used when locking the discussion thread on the change ([1] refers to [2]).
I think the burden of proof is on you to argue that Codeberg will stop at exactly two policies, given they haven’t said they will do that, but they have already banned things they don’t like. A slippery slope argument isn't a fallacy when you've already slipped down the slope twice.
I really don't understand the argument, to be honest. If somebody is participating in something with a perceived expectation, and then that expectation gets contradicted not once, but twice, why is it unreasonable to assume it might happen again? I'm not being facetious, I genuinely don't understand the line of thinking.
If they have a coherent set of values, then that's why. Then you have a theory of mind about the admins, and they weren't behaving erratically, but it was your original expectations that were incorrect.
Codeberg has existed since 2018. Was it clear to you a few years ago that things like 1) projects tied to AI, 2) mostly AI code, and 3) crypto would be banned? If not, then what makes you believe you now know their future actions if you didn’t before?
If you consider the argument of the comment at the top, that those who disagree with banning will leave. It would seem that you could expect an acceleration.
I have actually seen, from the inside, organisations implode though ideological purity demands.
No two people believe in all the same things. In the end this path reduces to one single shared value, That those with different values should be excluded.
They won't. No company will, I think the best quote I've heard this week is from the proton CEO on the Decoder podcast.
"No company is going to jail for you."
If vibe code is of questionable legality, then the best thing to do is to remain guarded until its legal standing is determined. It is simply an added bonus that it prevents slop that consumes data on servers for projects that nobody cares about enough to not have vibe coded.
Are you arguing a Codeberg member had a legitimate fear of being jailed for hosting others’ AI code? And GitHub/all other hosters should have the same legitimate fear of being jailed a long as they don’t ban the same? Or is this more of a straw man? Or were you bringing up that quote to point out that those trying to defend Codeberg are using completely fabricated arguments?
While I can't find something specific about it in the new ToS, the new blog post notes that "side projects and experiments", and "specific tools and custom scripts that would be unlikely to find a community anyway" are "discouraged", but are "tolerated in practice" if they only use minimal resources. This is stated as being regardless of LLM use. So some of my repositories are apparently already borderline in the community's opinion (it's not clear how they judge use of resources, but I only use self-hosted runners).
But there also seems to be an increasing sense that many active members see Codeberg as a space for a particular type of FOSS, within a particular, normative view of what development and communities should look like. This ties into the "side projects and experiments" comment, and to the general tone of the AI blog post on what they want Codeberg to be and not be. The point about "projects with a single developer and virtually no users", about "large ghost projects", and so on, are within the context of AI, but are derided in a wider sense.
Some of my repositories look like single developer experiments with no users or community, but aren't. I'm ultimately a scientist, not a typical developer; users typically communicate by email and at conferences, and cite my software and associated papers in their papers rather than starring repositories on forges.
And yes, much of the discussion around these questions often falls into the argument that projects that are good will be fine, because they'll be accepted despite the rules, and that any concern about specifics is unwarranted. But I can't know that I won't be targeted by Codeberg having an increasingly normative view of what 'proper' community development looks like.
(Yes, I realize I'm free to use something else. I'll likely end up moving my repositories off Codeberg and onto some combination of my own Forgejo instance and Github. The lack of federation in Forgejo heavily penalizes not being on a major instance, however. Meanwhile, as much as I encourage using DOIs for citations rather than URLs, people often do the latter, so staying on an instance that might decide to ban my projects can create lasting problems: one of my whole reasons for starting to migrate to Codeberg was because it seemed organizationally stable.)
My read of the "side projects and experiments" part of the blog post is that, in the context of LLM projects, if your project does not use resources and isn't looking to attract a community then, if only due to the practicalities of moderation/enforcement, they're unlikely to take any action against it because it won't be visible or causing enough problems.
Asking about this, it seems like that might be the case: they likely intended the "even if they were not LLM-generated" to be counterfactual. It's an unfortunate place to leave such grammatical ambiguity.
> You are effectively saying you do not trust the judgement of the community to accept you long term. Why is that?
I think the blog post explained exactly that. To quote:
> The actual wording makes this more difficult. The terms prohibit projects that mostly consist of code written by generative AI tools. In an actively developed codebase, what does “mostly” mean, and who can still tell? I could not reliably assign authorship percentages to many of my own recent projects. The line is open to interpretation precisely where it needs to be enforceable.
Hacker News recently banned LLM-generated comments due to a vague vibe check of the community. Do you feel the same way about the slippery slope of Hacker News moderation?
Did it get much attention? I don't know of a single person who finds value in LLM generated comments on social platforms. The fact that Codeberg is getting attention for this move is likely because it alienated a part of the community that was previously engaging, enough to generate constructive responses. I didn't see the same thing happening with this HN change.
1. dang specifically posted a comment saying the rule has been around for _years_ [1]
2. I browsed the top upvoted comments and they seem to almost universally lean towards this being a good thing.
That's about what I would of expected. Although the "being around for years" thing was news to me and doesn't really make the OP's argument a good comparison for Codeberg's case.
Yes, and the baseline amount of trust seems to be substantially reduced for Codeberg for a lot of people in the wake of this change relative to most of the other options. I don't even necessarily think the new policy is bad in a vacuum, but if the strongest argument for why you can trust a service is "well can you really trust anyone 100%?", it's hard not to be a little wary.
Would you use a bank if their response to the question of "How would I know I'll be able to access my money tomorrow?" is "This question exists for any bank" rather than actually giving any sort of reassurances?
It does exists for any bank. Where I'm at, the assurance is given by the national bank, not the individual banks. Anything else is just probability and marketing.
In Codeberg case, what I would be looking at is the actual banning process, not "assurance" they won't ban me later. It's a Germany-based institution, if Germany applies a whole sanction to my country, I'm gone anyway.
No one is disputing the existence of the question.
> In Codeberg case, what I would be looking at is the actual banning process, not "assurance" they won't ban me later.
Okay, so then, what actually is the banning process? The recent changes sound like the banning process has been expanded to cover a lot of new things that it didn't before, and stating "well anyone can ban you" isn't particularly helpful information, which was the point I was making.
> The recent changes sound like the banning process has been expanded to cover a lot of new things that it didn't before
This is the actual part of the TOS
Please be aware that this list is neither exhaustive nor complete (illustrating the spirit of the law, not its letter). Be aware that any account can get suspended at any time if it is used for activities harming Codeberg e.V. or its associated platform and services, directly or indirectly, even if these activities are not explicitly listed in the examples above.
So anything goes. The actual list contains stuff that are going to get you expulsed for sure. But there’s nothing that guarantees your ability to host for a certain period of time.
So your best bets is to keep backups of your data soemwhere else. Which you should always be doing.
> So your best bets is to keep backups of your data soemwhere else. Which you should always be doing.
Generally backing up a git repo is easy; it's all of the other stuff that tends to get hosted together with the code (issue tracking, pull requests, releases, potentially even CI) that's a lot worse to have to migrate over. Plus, anywhere a link was published to the repo is going to point to something that no longer exists.
It actually sounds like we're in agreement for the most part about the relative levels of risk here. I misunderstood your earlier reply to me as a rebuttal to what I was saying to the parent commenter, but from re-reading it I'm realizing I might have inferred incorrectly.
We are. My main point was that it's not worth it to disparage others because of their rule. Unless there's a contract or other legal clause, everyone is free to do whatever with their stuff. It's wholly a matter of trust. And I never trust any online service to be permanent in their behavior. I just enjoy it while it lasts.
> it's all of the other stuff that tends to get hosted together with the code (issue tracking, pull requests, releases, potentially even CI) that's a lot worse to have to migrate over. Plus, anywhere a link was published to the repo is going to point to something that no longer exists.
I think most of them provides API endpoint to access those data or some export tools. And as far as links go, I would prefer to have a webpage for any relevant projects, something like the one for msmtp [0], tied to my own domain.
Did you know GitHub removes things they deem to be malware? Malware is a very subjective classification, so who watches the watchmen? How do we know they won't start removing things that are deemed harmful just for competing with Microsoft products? Hell, letting them have a mechanism to moderate the site's content at all is a huge risk to society's freedom and continued existence. Aliens could arrive any time and shoot mind control beams at the GitHub admins which compel them to delete everything at once. The only way to prevent the future holocaust is if I'm allowed to typo squat GitHub repos to distribute viruses and someone else has to pay for it. If it's not safe for me to do that, GitHub is not safe for anyone.
And also like... this is how literally any community works? You "invest your free time" in it, then you do something that goes against the values of that community... yeah, of course that big bad "shadow group" (in other words, the admins/moderators) is going to do something.
It's not analogous at all. I moved my project to codeberg because the values were one way, and GitHub didn't match my values. I pour my heart and soul and time into my project. I also used AI tooling in a fashion as I saw fit. Now I'm reading I'm not welcome -- retroactively -- because I touched "bad" tools.
So I had to spend the last 3 days dealing with all the fall out from that and moving my project -- which was not insignificant -- docker images, debian package hosting, documentation references, existing in-flight pull requests, blog posts, build/release processes, my CI configuration, etc. etc. I'm nowhere near done.
An individual making an AI slop comment on HN loses that one comment to moderation and moves on.
I'm with you on this. I started moving projects to Codeberg when my understanding of the rules was that it was friendly to FOSS projects. Now it's explicitly unfriendly to a set of FOSS projects (albeit of a kind that I didn't participate in anyway) and users who use a particular kind of tool while working on their otherwise OK FOSS projects.
Now you can have a perfectly legal GPL-licensed project that's not welcome because of the category it finds into or because of its devs' tooling. It's their site and their rules, sure, fine, but it does mean I'm not going to host my policy-compliant projects there. Are they going to get banned next month? Who knows.
To be blunt, I think you may have misunderstood the values to begin with. A motion being passed by a majority is a lagging indicator to the values of the membership, so if they changed they will have changed a long time ago.
The Codeberg blog has articles critical of AI published back in September 2024[0]:
If they can use so much power to bring our systems down, imagine how much power they use by doing this to the entire Internet: We are concerned about the environmental and social impact of the current AI hype.
Growth at all costs is something that we, as a non-profit, are not interested in. We believe in frugal computing and that the resources our planet provides are limited. It is no coincidence that we directly "compete" against a company that funnels billions into the AI hype.
I don't even use Codeberg, and I've understood it was at best wary of AI for years. This shouldn't have come as a shock to you.
This is basically the argument I've seen over and over again from the Codeberg community. Boo hoo, you misunderstood the writing on the wall and now you're stuck doing a bunch of work. Goodbye.
I mean, more power to you (the community) in setting up your fiefdom, but by God do I suspect there's going to be some good people who leave, not wanting to participate in that sort of dialogue. I don't understand why it's so hard to show a little understanding and compassion.
The goal is to reduce the amount of use of the platform, and I think imposing this rule is an excellent way to do that. Not sure I'd call it genius, though. There's a fair likelihood of unintended consequences.
these things always end with community disintegration. Eventually, some high profile user will cross the undefined line or leaders change and want to rewrite the rule then half the community will side with them and half will not.
Yes, every community disintegrates. Eventually. Individuals also disintegrate. Over the long run, planets and stars disintegrate. Best not to think about it too much lest you go insane.
I think we should point out that Codeberg is a e.V. and that they have a Mission Statement [1]
> The mission of Codeberg e.V. is to build and maintain a free collaboration platform for creating, archiving, and preserving code and to document its development process.
As I understand it, they do not ban Crypto. Crypto is just an example for a category of projects that harm there reputation.
The formulation for AI generated code is different. It does not point out AI / LLMs as example for copyright issues. It does it the other way around.
This is a lesson I learned in life: When you want to build something social (in this case a community, but it could also be a romantic relationship, a friendship...), the best thing you can do is to polarize as quickly as possible.
Too many people are afraid to offend and waste everyone's time going along with things they don't really agree with and end up pretending to be somebody they're not to get along with people they don't really like.
No! Get your most polarizing takes out quickly. Either you and the other person stop wasting time because there's no fit, or you will deepen the relationship and both know it's a good fit. Either way, it's better for everyone involved.
There's enough people out in this big world that you don't need to always make everyone happy. It's much better to quickly comb through to find the gems.
> There's no issue with this. This isn't the supreme court. Your income is not on the line. It's a code hosting repository. Go to GitHub if it bothers you.
Codeberg is best known as a place for people to go when they're fed up with GitHub. People who are fine with GitHub's various issues wouldn't be looking at Codeberg in the first place, as far as I can tell.
There is GitLab, though. And Forgejo and a bunch of others.
The big thing I read is Codeberg is for community projects explicitly now, not just your random non-commercial code.
Personal use cases are now “might go unnoticed” and therefore not encouraged. Jeff’s dotfiles repo will probably never have a community, so it’s technically against the rules for some future cleanup initiative.
Most of my repos are random personal projects that are not supporting any larger open source thing, so I couldn’t look at (or recommend Codeberg) as a GitHub alternative for anything but an established project.
I think that’s all an acceptable choice, but it’s not being discussed much.
> The fact that the rule is open for interpretation means that they can effectively shut down malicious actors
This also mean that some kind of charismatic leader can become a de-facto tyrant.
And btw, lone leader on top, bending the rules at will… that’s a trait all dictatorships have in common (hitler, mussolini, stalin, mao, etc). Just saying.
This is the problem with choosing platform for moral — instead of practical — reasons.
Platforms that are known for making moral judgments will likely keep making moral judgments in the future. Eventually, they'll make a judgment that you heavily disagree with.
This is why communities founded on ideology tend to shatter, while those founded on a shared goal tend to survive. Linus doesn't care whether you're a GPU expert at AMD, a 16-year-old in Nairobi, or a nazi with a swastika tattooed on your arm. He cares about the quality of your kernel patches.
I think this shows that there is a blurry line between moral and practical judgment, and that the type of judgment might not even matter in the way you think. Sure, this episode may have been started by OFAC sanctioning various Russian companies, but Linus is happy to invoke more personal moral terms; and ultimately whether it's practical or moral doesn't matter, a bunch of Russian folks found themselves without a commit bit seemingly at random and the same could happen to anyone regardless of a particular platform's or company's judgment.
Is the fact that churches are some of the longest surviving communities accounted for?
(and of course Linus cares, like any unaccountable asshole he can be triggered by a way more trivial thing than a swastika (see the side comment for an example)
Perfectly summarised. Also recently saw so many people moving from GitHub to Codeberg because of anti-US and sovereign European tech movement and I wonder now which platform they are going to move again
They need to self host if they want sovereignty. This thing about relying on an external service and demanding total privacy and control is irrational.
Deciding that you "won't make moral judgments" is also a moral judgment. It's just a less obvious one. All platforms have made moral judgments in the past, and will continue to do so in the future. Some of them just don't realize it.
Its a choice they're totally allowed to make, I just wish we'd get an updated mission statement[1]:
"The mission of Codeberg e.V. is to build and maintain a free collaboration platform for creating, archiving, and preserving code and to document its development process."
Amended to include at the end ", as long as you use the tools we find agreeable."
I have no particular problem with this decision/policy on the part of Codeberg.
I do have an issue with their "tainted code" project that labels various FLOSS projects as "AI tainted".
ardour.org has the follow AI policy statement:
---------
Using LLMs to generate code for Ardour
A word about licensing, copyright and LLMs ("AI"). We require all code contributed to Ardour to be licensed under the GPL (you may retain copyright, but the code must use the GPL). Currently, following Thaler vs. Perlmutter in March 2025, no LLM-generated code may be copyrighted in the USA. The GPL relies on copyright, and cannot function if copyright is not held. Consequently, if you use an LLM (Claude, Codex, CoPilot etc. etc) to create code, it cannot be included in Ardour.
If you want to use LLMs to explore and understand some of Ardour's source code, feel free. But we cannot accept PRs or patches or contributions that are not licensed under the GPL, and if you choose to be honest about it, this means that no LLM-generated code can be accepted by us.
---------
Yet this codeberg-adjacent project considers this project to be AI-tainted. If you're going to do something, do it right or don't do it at all.
They didn't consult a lawyer when they wrote that. The idea that public domain code that is not copyrightable is incompatible with the GPL is simply not true. sqlite is in the public domain and not copyrightable, for example.
Their decision leans too heavily on what that court decision said. It was actually fairly nuanced.
The court explicitly declined to address whether Thaler could be considered the author because he created the AI, because Thaler originally listed the machine as the sole author in his registration application.
It is clear that works generated entirely by AI without meaningful human contribution are not eligible for copyright protection, though it does not resolve questions regarding AI-assisted works where human input is significant, which is a predominant use case with AI-assisted coding.
Finally, it appears they're unfamiliar with copyright rules around non-copyrightable works that are combined with other creative copyrightable works. The result is almost universally copyrightable. I'm hedging because I haven't thought through the cases exhaustively, but I can't think of a single case where the resulting work would not be copyrightable.
So they have modified their policy because they do not understand either copyright law in general or Thaler vs. Perlmutter in particular.
I think we will be seeing a lot more of this. There were plenty of examples of developers failing to understand copyright law and the implications of licenses before LLMs and that was with well established law. With some genuine uncertainty around LLMs, more variation between jurisdictions and some criteria that are less clear its going to get a lot worse.
> sqlite is in the public domain and not copyrightable, for example.
And therefore is not licensable under the GPL; because no license is needed at all. Ardour's goal isn't to be "compatible with" the GPL (in the sense of allowing binaries to be built containing both GPL-licensed code and their own code); their goal is to be covered by the GPL license for the purpose of constraining what actions users can take. This requires the code to be copyrighted, because copyright law is the mechanism that allows them to require users to agree to a license.
You're ignoring that adding non-copyrightable code to a collection of copyrightable code makes the collection copyrightable. The folks that wrote the policy are also ignoring that. It's simply a legal misunderstanding, because (I suspect) they didn't actually consult a lawyer, and their reading of precedent is overly broad.
Ardour does not seem to require a CLA or copyright assignment, so the maintainers don't own the copyright over contributions either way. The contributor would not be able to sue someone over a GPL violation if their contribution is public domain, but what random small contributor is going to participate in a GPL infringement suit anyway? Doesn't change the maintainer's rights to do so with any of their own human-authored contributions.
I don’t even mind forges taking a stance on AI. That’s fine. But I agree with Armin Ronacher here: banning projects that “mostly consist of code written by generative AI tools” is like the infamous Supreme Court “I know it when I see it” line. Yes of course I know a project mostly written by generative AI when I see it.
I don't understand why people are referring to Codeberg as competitors, or going to competition. They're WANTING you to go back => GitHub and consume the GitHub's resource's if you do not conform to their traditionalist nonprofit view.
They are not trying to retain you as a customer or even have customers ! They do not advertise ...! I am a [relatively responsible] utilisateur of AI and I truly do not understand the entitlement of some people.
It is a nonprofit, the vast majority of people complaining are not voters, donators, or members, while the vast majority of nonprofit members have chosen this view.
My personal opinion, is that the people clutching pearls are _scared_. There is a nagging feeling that deep down, they know the machine gods they love so dearly are _wrong_.
That if actions like these pick up momentum, if their "experiments" work, and more and more people start realizing that alternatives exist, the current places will end up as mostly "dead malls" as the people have moved on to the new places. And they will be facing at a gate where the people guarding it know where they came from, and why they are now asking for entry.
I think introducing unpredictable new prohibitions is a much bigger problem than the specific bans on cryptocurrency or AI written code. Especially when it leads to existing code becoming unavailable and doesn't just ban new contributions violating the rules.
They weren't exactly unpredictable. People didn't predict it because they're used to dealing with faceless machine platforms instead of human platforms. GitHub bans projects rarely, and when it does, it's without any warning or rule change.
Codeberg should host code, and should not be bothered by which tool the code is written. Let's wait until GitHub starts accepting only code written in VS Code or Visual Studio because those are Microsofts tools.
A notable surprise with Bluesky over the past few months is that the anti-AI sentiment has decreased slightly as it becomes harder to argue against impactful use cases.
At the start bluesky was actually pretty nice, but now it's just this weird bubble where disagreements don't exist and when I checked out like a month ago there was just so much damn p***.
Aren't there good filters for that? It's one of their core differentiating features - the ability to use custom content filters. Or did you check with no protection enabled?
This sounds great. We've hit some kind of scaling problem where "algorithms" are not up to the task of finding worthwhile stuff among the solar masses of junk being pumped out daily. And even if they were, the amount of exploitation is so fathomless that I don't think it could ever work.
It's time for society to rediscover the power and immense value of real curation.
Even with non experts, just looking through another human's playlist of recommended songs beats the crap out of algorithms. I've found so much more music that way than with youtube suggestions or whatever spotify does.
This decision may also be related to infrastructure. Even if Codeberg wasn't against heavy LLM use for moral reasons, their limited infrastructure would have a hard time keeping up with the massive output that LLMs are capable of.
For once I wish they would've taken a more technological solution to this social problem and enforced more limited CI time and disk space quotas by default. Seems like it'd be fairer and require less active policing. My read from their post was that the resolution had more of an ideological bent than being just about resources, which is certainly their community's right to have.
I am a heavy viber that uses a lot of resources in github now after LLMs compared with pre-LLMs for my personal projects.
I find great that there are places that are still human first and sensible about community and free software instead of just providing technical infra. Gitlab and Gitea are already open Github alternatives: open source core projects turned open sourcish business with free as in beer plans that get monetized with user data, loss leadering, business offerings and other sassy business models. Codeberg is deciding to not be that. Other projects can decide different and its users should evaluate the tradeoffs and choose.
> I could not reliably assign authorship percentages to many of my own recent projects.
This feels like a strange statement. Unless you are going out of you way to disguise AI contributions it's extremely simple to do some git magic to get an exact percentage. But you can also just ask AI and it'll tell you directly. And even if you've tried to disguise it I'm fairly certain that if you set aside 10$ for tokens, it'll be able to deduce how much is AI even with someone having tried to disguise it.
My only complaint about their site is that it mentions a generous free tier, but I have no way to seeing what exactly that free tier is.
Edit: Found it, but it wasn't obvious the navigation path to it. Would be nice if "free tier" on their home page was a link where it was there and clear.
> It can choose to be a smaller community with a stronger political identity, but that is a different ambition from being a broad and dependable European alternative to GitHub.
Wrong category. It doesn't want to. It's not a company. Might as well act on their beliefs.
Devs are so spoiled with free CI, running nonsensical workflows all the time for free. Imagine that multiplied by magnitudes thanks to AI. Just a waste of resources.
The facts are missing in the Article. The changes really made by Codeberg e.V.
are:
First:
> 7. You must not share projects that mostly consist of code written by "generative AI"-tools (including services such as Claude, OpenAI Codex). Such projects having an unclear copyright status (see requirements § 2 (1) 1 and § 2 (1) 3) and furthermore have little safeguards to ensure that they do not include harmful code (c.f. § 2 (1) 5).
Second:
> - Content that harms the reputation of Codeberg, such as cryptocurrency related projects.
Codeberg was competition for the EU, in the topic of tech-sovereignty which is a incredibly hot topic in the EU. With the US abusing it's position not only recently but years past, the EU is accelerating funding and plans to decouple from anything the US can mess with legally. Much of that sentiment has been spreading all over the EU governments and software communities.
GitHub for all its purposes remains a platform hostile to EU interests long term.
I don't think they were saying that there aren't reasons to not use github, in or out of the USA, but rather that codeberg was already not a contender for many use cases due to their existing rules restricting private/non-OSS repos.
> I wish Codeberg were more forward-looking here: willing to host the Open Source software of tomorrow, not only software made in the ways its community approves of today.
Is "vibe-coded software" really going to be the "Open Source software of tomorrow"? The opinion stated in the blog post announcing this change is the opposite:
"We believe that LLMs endanger the free/libre software ecosystem as a whole." and this is expanded in more detail later in that blog post (starting at the "Collaboration at danger" section).
This reeks of the same FOMO strategy seen often in cryptocurrency projects (which, unsurprisingly, have also been banned from Codeberg at the same time): stating that their way is going to be the future, and that those who don't follow it will be left behind ("have fun staying poor").
> But these tools are also becoming part of how software is made. The Open Source world needs to figure out how to engage with that future, not just divide into camps.
No. They don't. This isn't a "you just have to deal". You can say no. It's OK to decide that it's not worth the smaller amount of positives to deal with the massive amount of negatives. If you disagree, start your own community.
There’s a lot of hue and cry over what’s just Yet Another Bulletin Board because it has a git repo attached? Just host this stuff yourself guys. It’s trivial these days.
I have my own remotes with my own CI and it’s all ultrafast because I can just make assumptions about the way I write software.
None of these platforms are necessary for development. Just take from them what you want and build your own universe. This grade of CRUD software is generally solved.
There are places you need other people. Stuff like ffmpeg or GPU kernels. But other kinds of code is kind of unimportant. Not being part of this community loses you nothing.
Is there any evidence of division? Are all the vibe-coders going to some codeberg2? One of the main complaints is that the banned projects don't have human users proportionate to their activity, so if they all go to codeberg2, then that forge would just be full of projects nobody cares about.
Are there any projects on codeberg that are getting banned and actually have significant human interest?
If codeberg decided to kick out spammers, nobody would decry the loss, so what's being lost here? Is it strictly slippery slope?
Nobody has been banned yet, AFAIK. So nobody can answer your second paragraph question. But plenty are choosing to leave.
The judgement you're talking about here ("banned projects don't have human users propostionate...") is going to require evidence? At this point the software community is absolutely full of these tools and people who use them.
If the specific thing banned was projects like that, fine. But that's not what the clause added in the PR said.
Maybe it will push more people and more effort into the true decentralization since the Forgejo/Codeberg federation[1] effort is way too slowly moving. Radicle[2] it is.
Just self host if you want freedom. I switched to Codeberg to maintain a semblance of project discoverability while having freedom from Microsoft but I don’t need the code police telling me what I can build or how I can build it. My open source values are more WTFPL "freedom to do whatever you want" and Codeberg is more AGPL "freedom to do whatever we allow you to", which is totally fine.
It seems to me, Codeberg sees no value in AI slop and doesn't want to spend significant amounts of its limited resources hosting it.
The author talks a lot about what they want from Codeberg. And I suppose it's fine to wish someone (on the right side of the ocean) would provide services, tailored to your personal desires, for free. But don't be surprised when no one takes you up on it.
For myself, I don't see a point to Codeberg simply being a clone of GitHub. GitHub is free, which is great, but it's also pretty shitty. Why do we need more of that?
I see. We should not divide into camps so the Austrians can sell their AI startups.
The only people dividing and ruining open source are those paid for it by AI companies. Sometimes indirectly, like NumPy through its various sponsors. Then suddenly a bureaucrat appears on a mailing list or elsewhere and announces that AI contributions need to be allowed.
Fully agree. I am against AI slop, LLM overhyping, and cryto scams, as much as anybody, however:
> the main property I need from infrastructure [is for] it to be predictable, dependable, and reasonably neutral towards the legal Open Source software hosted on it.
I do not like to build on a foundation that could be ripped away from under my feet at any moment for arbitrary reasons.
> I do not like to build on a foundation that could be ripped away from under my feet at any moment for arbitrary reasons.
Do you only build and publish software in a vacuum running on hardware you own? There is no online community in the western hemisphere that isn't subject to any kind of arbitrariness, be it internal (community, like here) or external (legal or infrastructurally, like on github).
I reiterate what I said in another thread (and invoke Don Mattrick): If you want AI slop, there's a perfectly fine*** product for you already out there - it's called GitHub.
For those of us who are tired of AI messing with things, stop complaining and let this be. Codeberg made a great decision for those who are just tired and over AI coding unnecessarily taking over everything.
> The Open Source world needs to figure out how to engage with that future
How ironic. Why is the burden of labor on the open source community? They have to do more volunteer work for your AI satisfaction, even after training the very models you depend on? I think proponents of AI are the ones who have the burden of figuring out how to engage with the open source community. Use your own tokens to figure it out. Surely AI is up to the task. Your frontier models have the entirety of human diplomacy at its fingertips to steer you towards the path of success. I wish you luck.
> More importantly, LLMs if done and used well, should be welcome to all of us. They could be used to reclaim control and power, away from large corporations and institutions.
This decision is consistent with the community Codeberg has attracted of late. It's smaller, independent and takes stances that are not in line with competitors like GitHub. If you want something less reliable, at larger scale and filled with AI projects that GitHub is clearly encouraging, you should use GitHub.
Genuine question to AI users upset by this: if you trust AI to generate your code, why don't you trust it to help you "deploy https://forgejo.org/ on cheap VPS make no mistakes"?
on April 1 next year don't be surprised when lots of repositories come out with "surprise! i was AI all along". How are they ever going to know? Seems like a prompt like "remove all traces from this code that it was generated using AI" would one shot the whole repo and make it virtually impossible to justifiably claim it was vibecoded.
> There are serious questions about copyright, labor, energy use, slop, and maintainers drowning in generated contributions. But these tools are also becoming part of how software is made.
This first sentence is such a weasel-acknowledgement that is used to just effectively ignore all of it.
Agree substantially here, as I said elsewhere: the honest and respectful answer from the Codeberg community would have been: add resource limits, quotas, and a reviewed application form for new projects.
And maybe a statement that: Pre-existing clearly "vibe coded" projects providing little value to the community may be subject to review. Our resources are limited. Don't abuse them.
Instead it got framed in moralistic terms in many places, the actual voting process actively resisted all calls for clarification ("if you're a vibe coder you'll know, so stop trying to ask for clarification", the tone in online forums became nasty, and the result is both ambiguous and culturally poisoned and I decided to move my project ("good riddance!" call out the victors, "we got one of them!")
I do think the one substantial argument that was made through the threads, was around IP ownership laws. However I remain unconvinced.
Frankly if there's anywhere where I would most like to see AI generated work -- which is frankly based on weights generated from the broad and indiscriminate harvesting of other people's works -- is in copylefted software that takes the distillations of the efforts of the community and makes it available as free software with free software protections.
I understand not everyone shares that opinion, and they're welcome to different interpretations.
Like the author of the article I'm saddened to see this split. I am honestly curious to see where we are in 2-3 years on this topic.
> Frankly if there's anywhere where I would most like to see AI generated work -- which is frankly based on weights generated from the broad and indiscriminate harvesting of other people's works -- is in copylefted software that takes the distillations of the efforts of the community and makes it available as free software with free software protections.
I agree wholeheartedly with this. We have large corporations stealing work left right and center. Lets utilise that to our benefit and make open and free versions of all software. The crux is that to do that each individual has to pay currency to use this tool. Maybe there is a place for collectives to collect money and to utilise llms to free software. (What strange times we live in.)
A slight aside, open source lived and breathed the idea of ownership of software, be that a council, a handful of committed maintainers, a BDFL and all things in between and adjacent. There isn't really much wrong with "vibe coded" software, the issue seems to be owning that software. Things feels fundamentally more fluid, based on specs and more importantly tests that actually define the features that the software is expected to perform. Not ownership, but more like a contract.
A great way to ensure that most developers will not be using Codeberg going forward. While there's a ton of drama around LLM use right now, anybody who's actually used this tech in anger understands that these tools really do work, and that they will only be used more in the future. I guess Codeberg is angling to become a small artisanal community, and there might certainly be a niche for that.
The outcome is obviously going to be that most coders leave Codeberg because LLMs are not going anywhere, and they will continue to be integrating into people's workflows. But perhaps Codeberg intends on becoming something akin to a digital Amish community, and there's nothing wrong with that.
"Not going anywhere" is such a strange argument. Tomato juice isn't going anywhere, does it mean I'm going to be guzzling it at some point in my life? What?
If you've made your choice about LLMs based on morals or ethics rather than the quality of the tooling, you're not going to change that choice because the models haven't disappeared or have improved, unless you're forced to (but that forced bit applies only to commercial open source which really just isn't on Codeberg at all in the first place).
It has been long enough that just about everyone has made that choice by now. Those who are okay with this won't magically flip and start using LLMs and grow a reason to leave.
Every moral or ethical argument ultimately boils down to problems inherent in how any piece of automation is used under capitalism, and has little to do with technology itself. If you spend a bit of time thinking about morals or ethics you'll quickly realize that LLMs are not the problem.
And no, a year is absolutely not long enough for everyone to have made a choice about this tooling. Claiming any sorts of absolutes about completely new technology that's rapidly evolving is absolutely delusional.
It sure is! But I guess it would be less fun to say "1 out of 5 developers will not be using Codeberg going forward". Especially since that may well be an intended effect of Codeberg's decision here.
As long as AI is a source of environmental damage, cognitive harm, and a cause of rising hardware costs, there will always be a sizeable set of developers who want nothing to do with it. The headline should be "AI Divides".
There's no basis for thinking that LLM use will not continue to grow going forward. The fact that 1 in 5 developers already use this technology only a year or so after it's become really useful shows that adoption is growing rapidly.
Meanwhile, the arguments regarding environmental damage and cognitive harm are rather spurious. You can already run models like Qwen 3.6 27b on your laptop today, and it can do serious work. And the efficiency of the models only keeps improving. Meanwhile, studies showing cognitive effects you allude to have used laughably small samples, and these same types of arguments were made every time the way people work with code changed. I'm old enough to remember people screeching how you weren't a real programmer if you didn't use assembly, or that GC was ruining programming.
There's always a group of people who made a certain set of skills their whole identity who get really upset when those skill become automated. There will always be a sizeable set of current generation of developers who will want nothing to do with it, but people getting into development today will treat LLMs as a part of their regular toolkit. At the end of the day it's just a tool like any other.
> There are serious questions about copyright, labor, energy use, slop, and maintainers drowning in generated contributions. But these tools are also becoming part of how software is made. The Open Source world needs to figure out how to engage with that future, not just divide into camps.
Everyone who says stuff like this is just burying their head in the sand. That is like saying that there are serious questions about the ethics of buying sweatshop-made clothes, or driving a gas-powered car, or robbing grocery stores --- but those are becoming part of life, and so we need to "figure out how to engage with them".
That doesn't mean you have to upend your life to fight the good fight. You can say "I want a world without X, but right now I need X to live, so I do X, while maintaining an awareness of it and looking for opportunities to do less X". You can say "I want a world with some kinds of X but not others, so I do the X that I'm okay with". You can even say "I think X is ethically fine so I'll do it". But it just doesn't make sense to say "there are ethical issues with X, but those only affect how we engage with it, and have no bearing on whether we think people should need to engage with it at all".
There can't be ethical consideration of "how" without at least considering the ethics of "whether".
> There are serious questions about copyright, labor, energy use, slop, and maintainers drowning in generated contributions. But
Way to not at all engage with any of that and just go on your merry way...
Sure, if you fail to acknowledge that the other side has valid points, it is easy to act like you're completely right and they are morons, and we all need to Kumbaya together, and forget about their views.
The way I read the article was not that the author disagrees with the concerns you mention, but instead that they believe Codeberg's recent actions fail to address these underlying concerns while adding uncertainty for users of their platform (which they fear will ultimately hurt codeberg). To me, it reads as a call for further discussion and clarification rather than a dismissal.
I think this move by Codeberg is lowkey genius. They're trying to build a community based on a set of shared values. Those who disagree with those values are now self-selecting out of the Codeberg community, which is exactly the point of passing these kinds of resolutions.
The fact that the rule is open for interpretation means that they can effectively shut down malicious actors (i.e those who intentionally misrepresent the degree of AI assistance in writing their code) before they become a problem. It becomes an exercise in judgement instead of a potentially exploitable/loophole-able set of hard and fast rules.
There's no issue with this. This isn't the supreme court. Your income is not on the line. It's a code hosting repository. Go to GitHub if it bothers you.
Only time will tell if this set of values holds up. Personally, I think they will, and that the rules will end up being less strict than people are making them out to be, and that relatively few projects will be removed from the platform.
I’ve seen so many variations of this where a community starts adding more rules that are open to judgment.
Removing the obvious rule breakers looks like an easy win at first. Then over time the list of rules grows. Small groups become good at navigating the rule making process and getting changes pushed through. Before you know it, small groups of highly invested people are wielding the rules and the rule making process against their enemies or competitors. The community was long since cleared out from anyone who got sick of the drama, leaving a weird internal battle where different in groups are fighting with each other over little things and turning the rules against each other.
Maybe Codeberg will be different. Maybe the rule making stops here and it’s reasonably enforced against only the really bad examples, with leeway granted to others generously. Maybe it doesn’t turn into a slippery slope of little dramas that exhaust everyone who isn’t invested in the constant background drama of their code hosting site. I hope it does for their sake.
But I also think enough people like me have seen what happens to communities that go down the path of generating a lot of rules like these and will choose to steer clear of this site. It always feels like a good idea when you and all your friends are on the right side of the rules, but it’s less fun when you find yourself making choices or tip toeing around to avoid triggering the Nth round of rules that were introduced 10 steps past where you thought the rules were in a good and balanced place.
>Before you know it, small groups of highly invested people are wielding the rules and the rule making process against their enemies or competitors.
Now that you mention it, Wikipedia also has a policy against LLM contributions.
Importantly, Wikipedia's policy is a bit more detailed about what LLM contributions are and aren't allowed. There is always ambiguity, but going through specific cases and examples up front in the policy went a long way.
How do you propose the people who want a shared space where they can know that there is no slop/crypto crap approach it?
Not to mention what happens when these rules cease to be rules on purely technical matters. What if they make a rule against the site allowing hosting the code *of* AI agents for example?
Codeberg did make such a rule. Their announcement explicitly says that "Projects heavily tied to the LLM ecosytem" are "no longer [...] welcome on Codeberg". The code for an AI agent would definitely fall into that category.
That's not true. The new "rule" say:
> 7. You must not share projects that mostly consist of code written by "generative AI"-tools (including services such as Claude, OpenAI Codex). Such projects having an unclear copyright status (see requirements § 2 (1) 1 and § 2 (1) 3) and furthermore have little safeguards to ensure that they do not include harmful code (c.f. § 2 (1) 5).
https://codeberg.org/Codeberg/org/commit/71149c7fc95ccfeae36...
Yes, that is unrelated to the rule he is refering to.
Then you have your agent create you your own codeberg, with blackjack, and hookers.
Can the LLM also materialise a server and free electricity for me?
No, but it can pretend to do that, which is close enough.
You're paying for the LLM already, which is running on a server and consumes electricity.
So you want free hosting and free services and you also want to dictate how you are able to use them?
Are you of the opinion that you are entitled to a free server and electricity?
Why would you need a server? a pi would do.
A Pi running a server is also a server
I've been thinking along these lines more and more. Trying to specify exact limits and trying to account for every possible situation when making rules just encourages people to find holes and workarounds. It creates a game, a competition between the rule makers and everyone else.
When instead you state higher level values and ask for cooperation, it encourages us to work together. It's the Honor System instead of a deliberate competition between rule makers and everyone else.
Concrete example: sports
When you are playing with a group of friends and there's no referee, most everyone does their best to just play hard and follow the rules. It's not perfect, but it works for the most part because we all love the game and want to have a good time.
Once you introduce referees, the game within the game of "what can I get away with?" starts and you end up with players learning how to foul without getting caught and then players overreacting to fouls in order to make sure the rule breakers still get caught, and then it evolves into flopping to get a foul called even when there was none, and eventually people start to hate the sport altogether (see: NBA, FIFA)
Concrete example: government and police forces. Do I even need to elaborate on this one?
That's actually how EU law works! I think it's also why US companies keep getting in trouble here, they try to loophole-lawyer like at home, but it just... doesn't work that way.
I had a whole long conversation about this with another commenter a while ago if you're interested in learning more: https://news.ycombinator.com/item?id=48707697
Very interesting discussion, thanks for sharing that.
India has a different system where generally speaking, it behaves more like UK/US courts (common law) over civil law.
But if Supreme Court reads fundamental rights are violated then Supreme Court can read the constitution in a similar way but Supreme Court has used it sometimes to force the government hands.
For example, when the Bhopal Gas tragedy happened where people died because of the accident (Rest in peace), instead of having individual people try to sue the govt. and fight for years with no end in sight, the supreme court used article 142 to impose single lump sum settlement and centralized compensation scheme. So it can force a govt to do something.
The Sahara-SEBI Case is even more direct when supreme court forced a company to refund amount and fast-tracked it in a more civil-law style case.
But Indian Supreme court is very slow and from my understanding, does try to use it rarely and only the Indian Supreme court can invoke article 142 and its more so done to prevent genuine gaps rather than already existing provisions
I found this whole discussion really interesting
In theory this split between common law and civil law is about who you trust more: judiciary or legislature. Civil law leaves more to interpretation for judiciary whereas common law leaves very little to interpretation. There doesn't seem to be very wide consensus on which system is necessarily better to be honest.
> there's no referee, most everyone does their best to just play hard and follow the rules
Only in small groups. Any sufficiently large groups will have a number of people (possibly sociopaths) that will break the rules for personal gain. And in a lot of cases it's enough if very few people do that.
That's where gatekeeping comes in.
That's right, and that's when the rest of the people say "we're kicking you out", and if the offender says "but why" everyone say "because you broke the rules" and if the offender tries to rules-lawyer everyone says "it doesn't matter, we're kicking you out because we believe you are not playing by the rules".
Based on this precedence, if I'm accepted into the community today, how would I know I'll be accepted tomorrow? If you're literally just using Codeberg as a mirror, then sure, it doesn't really matter much. But like you said, that's not what they are building. They want people to invest in the community. So now, as an investor of my hard-earned free time, I'm faced with the question on whether all my investment can go up in flames on the whims of a shadow group.
You are effectively saying you do not trust the judgement of the community to accept you long term.
Why is that? Is it because you believe there would be a new policy that would exclude your workflow? What would such a policy be? Do you really, reasonably believe it would be implemented "tomorrow"? Do you think adapting to such a policy would be unreasonable?
Or is it because you don't trust the judgement of the people enforcing these new rules?
If not one of the two, what is the reason?
The point I'm trying to make here is that this just sounds like a slippery slope argument. If you don't have a concrete reason to believe you won't be welcome "tomorrow" then my mind has not been changed.
> You are effectively saying you do not trust the judgement of the community to accept you long term.
Correct. I was previously engaging with this community and had moved some repos over from GitHub. One of those repos deals with cryptography that has an intersection with "crypto" projects. Due to vagueness of the language ratified, it's unclear whether the repo now violates the ToS. This completely caught me off guard, and now my trust in the judgement of the Codeberg community has been weakened.
This feels like a reasonable reaction to me.
So it sounds like you're treating Codeberg as a neutral git host. They've made it clear they don't want to be that, they are a charity association that supports free software development by providing free software developers with services like git hosting. They are not wasting their services on people who don't align with the mission.
That seems like a completely different concern than what you originally stated? You're not worried about future policies, you're worried about the policy that was just passed.
And for what it's worth, it seems like they only care if the project is directly crypto related (e.g if you had a repo directly implementing a Monero node, that would probably not be allowed). If it's just a repo that implements a generic Cryptography concept that happens to be used both in crypto and elsewhere, then I don't see what the problem would be.
Again, the rules are judgement based. That means you can interact with the community to get clarification, and make your case that your project is worth keeping.
> That seems like a completely different concern than what you originally stated?
I started with a hypothetical. Then I used an actual real-world example. I have 100+ repos on GitHub spanning a wide array of technologies. If I moved all 100+ over, how do I know I won't fall afoul of whatever the community votes to ban next?
The point is Codeberg isn't a good fit for my situation which needs more continuity and less surprises, and that's been made readily apparent now. So I'm leaving, just like many other people who have been posting about it.
Luckily for you, as far as I can tell you don't have any concrete reason to fear having your repos taken down. Or at least I haven't been given any.
If the reason is more emotional than rational ("this feels like it's going to change more"), that's okay. Not everything needs to be governed by rigid rules. The Codeberg community knows that well. If in a few years your 100+ repos are still all fine and dandy to their rule set I'm sure they'd be happy to have you.
> Luckily for you, as far as I can tell you don't have any concrete reason to fear having your repos taken down. Or at least I haven't been given any.
Unfortunately you can't make that decision. I have to wait and find out how far the ToS violation will be implemented. Or I just avoid that stress and go somewhere else. Most people being affected seem to be choosing the latter.
> If the reason is more emotional than rational
That's a bit reductive. Risk management always has a level of uncertainty involved. No one makes decisions with 100% certainty in the outcome. Before these votes, I had a certain level of risk assessment in moving off Github To Codeberg that favored Codeberg. Now that situation has changed. It's not "emotional" to recognize that.
Have you considered actually asking someone at Codeberg about your cryptography project?
Who would one ask to get an official and enforceable answer?
I don’t want an opinion, I need a definitive ruling I can trust.
It's not like they don't have a contact page: https://docs.codeberg.org/contact/
> I need a definitive ruling I can trust
Why? What's so special about the place where you host your source? It's the second time I see this sentiment in the thread, and I don't get it. If an official channel tells you a project is OK, and then someone else decides it's not, what does it really happen outside of you taking your project elsewhere?
Hosting a project with its issues, packages, releases, discussions and everything is an investment. The URLs will be in multiple places, some of which I don't control.
Then suddenly the hoster decides that my project isn't wanted because the zeitgeist is against some technology used in it and poof it goes away.
What now?
"Just move away", sure. I can do that if they still let me export my data. But to be honest I'd rather just use a platform with a smaller risk of getting booted because someone decided my project is "too AI" or something similar.
I have no idea if there's a way to argue against the decisions, what proof do I need to provide, do I need to prepare somehow? Like have video of myself with a dated newspaper typing the code by hand to prove AI wasn't used? :D
Ok then do that. They're not asking for your business. They lose money on it.
Is this the "contact our sales team for pricing" of open source?
no this is the "we are real people, not a faceless algorithm" of open source.
>And for what it's worth, it seems like they only care if the project is directly crypto related (e.g if you had a repo directly implementing a Monero node, that would probably not be allowed).
Looking at community conversations, there actually seems to be little clarity over this, and quite a bit of confusion. Several established members of the community seem to be saying that the ToS actually only bans "cryptocurrency related projects" that "[harm] the reputation of Codeberg", and that direct cryptocurrency projects that don't harm Codeberg's reputation are actually allowed, saying that the current banner is a mistake and is in the process of being changed (while they have their own systems and are not on Codeberg, I'm guessing GNU Taler might be an example of a cryptocurrency project that would be allowed). In fact, this appears to be within the official interpretation used when locking the discussion thread on the change ([1] refers to [2]).
[1] https://codeberg.org/Codeberg/org/pulls/1254#issuecomment-19... [2] https://codeberg.org/Codeberg/org/pulls/1254#issuecomment-19...
I think the burden of proof is on you to argue that Codeberg will stop at exactly two policies, given they haven’t said they will do that, but they have already banned things they don’t like. A slippery slope argument isn't a fallacy when you've already slipped down the slope twice.
I really don't understand the argument, to be honest. If somebody is participating in something with a perceived expectation, and then that expectation gets contradicted not once, but twice, why is it unreasonable to assume it might happen again? I'm not being facetious, I genuinely don't understand the line of thinking.
I think I agree with you, so maybe you meant to ask someone else? Or you’re just supporting my argument, in which case I appreciate it :-)
If they have a coherent set of values, then that's why. Then you have a theory of mind about the admins, and they weren't behaving erratically, but it was your original expectations that were incorrect.
Codeberg has existed since 2018. Was it clear to you a few years ago that things like 1) projects tied to AI, 2) mostly AI code, and 3) crypto would be banned? If not, then what makes you believe you now know their future actions if you didn’t before?
If you consider the argument of the comment at the top, that those who disagree with banning will leave. It would seem that you could expect an acceleration.
I have actually seen, from the inside, organisations implode though ideological purity demands.
No two people believe in all the same things. In the end this path reduces to one single shared value, That those with different values should be excluded.
They won't. No company will, I think the best quote I've heard this week is from the proton CEO on the Decoder podcast.
"No company is going to jail for you."
If vibe code is of questionable legality, then the best thing to do is to remain guarded until its legal standing is determined. It is simply an added bonus that it prevents slop that consumes data on servers for projects that nobody cares about enough to not have vibe coded.
Codeberg is not a company. Also, we're not talking about jail. We're talking about principles.
> No company is going to jail for you.
Are you arguing a Codeberg member had a legitimate fear of being jailed for hosting others’ AI code? And GitHub/all other hosters should have the same legitimate fear of being jailed a long as they don’t ban the same? Or is this more of a straw man? Or were you bringing up that quote to point out that those trying to defend Codeberg are using completely fabricated arguments?
Honestly, for me, both.
While I can't find something specific about it in the new ToS, the new blog post notes that "side projects and experiments", and "specific tools and custom scripts that would be unlikely to find a community anyway" are "discouraged", but are "tolerated in practice" if they only use minimal resources. This is stated as being regardless of LLM use. So some of my repositories are apparently already borderline in the community's opinion (it's not clear how they judge use of resources, but I only use self-hosted runners).
But there also seems to be an increasing sense that many active members see Codeberg as a space for a particular type of FOSS, within a particular, normative view of what development and communities should look like. This ties into the "side projects and experiments" comment, and to the general tone of the AI blog post on what they want Codeberg to be and not be. The point about "projects with a single developer and virtually no users", about "large ghost projects", and so on, are within the context of AI, but are derided in a wider sense.
Some of my repositories look like single developer experiments with no users or community, but aren't. I'm ultimately a scientist, not a typical developer; users typically communicate by email and at conferences, and cite my software and associated papers in their papers rather than starring repositories on forges.
And yes, much of the discussion around these questions often falls into the argument that projects that are good will be fine, because they'll be accepted despite the rules, and that any concern about specifics is unwarranted. But I can't know that I won't be targeted by Codeberg having an increasingly normative view of what 'proper' community development looks like.
(Yes, I realize I'm free to use something else. I'll likely end up moving my repositories off Codeberg and onto some combination of my own Forgejo instance and Github. The lack of federation in Forgejo heavily penalizes not being on a major instance, however. Meanwhile, as much as I encourage using DOIs for citations rather than URLs, people often do the latter, so staying on an instance that might decide to ban my projects can create lasting problems: one of my whole reasons for starting to migrate to Codeberg was because it seemed organizationally stable.)
My read of the "side projects and experiments" part of the blog post is that, in the context of LLM projects, if your project does not use resources and isn't looking to attract a community then, if only due to the practicalities of moderation/enforcement, they're unlikely to take any action against it because it won't be visible or causing enough problems.
Asking about this, it seems like that might be the case: they likely intended the "even if they were not LLM-generated" to be counterfactual. It's an unfortunate place to leave such grammatical ambiguity.
Is it not obvious? They've just done it. They've excluded people based on new rules, and not everybody was expecting those new rules.
Having done it once, they can clearly do it again. Will it be justified? Who knows? Will it exclude you? Maybe.
> You are effectively saying you do not trust the judgement of the community to accept you long term. Why is that?
I think the blog post explained exactly that. To quote:
> The actual wording makes this more difficult. The terms prohibit projects that mostly consist of code written by generative AI tools. In an actively developed codebase, what does “mostly” mean, and who can still tell? I could not reliably assign authorship percentages to many of my own recent projects. The line is open to interpretation precisely where it needs to be enforceable.
Hacker News recently banned LLM-generated comments due to a vague vibe check of the community. Do you feel the same way about the slippery slope of Hacker News moderation?
Did it get much attention? I don't know of a single person who finds value in LLM generated comments on social platforms. The fact that Codeberg is getting attention for this move is likely because it alienated a part of the community that was previously engaging, enough to generate constructive responses. I didn't see the same thing happening with this HN change.
> "Did it get much attention?
4229 points and 1716 comments at https://news.ycombinator.com/item?id=47340079
Interesting. Two things stand out to me:
1. dang specifically posted a comment saying the rule has been around for _years_ [1]
2. I browsed the top upvoted comments and they seem to almost universally lean towards this being a good thing.
That's about what I would of expected. Although the "being around for years" thing was news to me and doesn't really make the OP's argument a good comparison for Codeberg's case.
[1]: https://news.ycombinator.com/item?id=47341726
> how would I know I'll be accepted tomorrow
This question exists for any service. Unless you host things yourself, there's a baseline amount of trust you have to have in any service.
Yes, and the baseline amount of trust seems to be substantially reduced for Codeberg for a lot of people in the wake of this change relative to most of the other options. I don't even necessarily think the new policy is bad in a vacuum, but if the strongest argument for why you can trust a service is "well can you really trust anyone 100%?", it's hard not to be a little wary.
Would you use a bank if their response to the question of "How would I know I'll be able to access my money tomorrow?" is "This question exists for any bank" rather than actually giving any sort of reassurances?
It does exists for any bank. Where I'm at, the assurance is given by the national bank, not the individual banks. Anything else is just probability and marketing.
In Codeberg case, what I would be looking at is the actual banning process, not "assurance" they won't ban me later. It's a Germany-based institution, if Germany applies a whole sanction to my country, I'm gone anyway.
> It does exists for any bank.
No one is disputing the existence of the question.
> In Codeberg case, what I would be looking at is the actual banning process, not "assurance" they won't ban me later.
Okay, so then, what actually is the banning process? The recent changes sound like the banning process has been expanded to cover a lot of new things that it didn't before, and stating "well anyone can ban you" isn't particularly helpful information, which was the point I was making.
> The recent changes sound like the banning process has been expanded to cover a lot of new things that it didn't before
This is the actual part of the TOS
So anything goes. The actual list contains stuff that are going to get you expulsed for sure. But there’s nothing that guarantees your ability to host for a certain period of time.So your best bets is to keep backups of your data soemwhere else. Which you should always be doing.
> So your best bets is to keep backups of your data soemwhere else. Which you should always be doing.
Generally backing up a git repo is easy; it's all of the other stuff that tends to get hosted together with the code (issue tracking, pull requests, releases, potentially even CI) that's a lot worse to have to migrate over. Plus, anywhere a link was published to the repo is going to point to something that no longer exists.
It actually sounds like we're in agreement for the most part about the relative levels of risk here. I misunderstood your earlier reply to me as a rebuttal to what I was saying to the parent commenter, but from re-reading it I'm realizing I might have inferred incorrectly.
We are. My main point was that it's not worth it to disparage others because of their rule. Unless there's a contract or other legal clause, everyone is free to do whatever with their stuff. It's wholly a matter of trust. And I never trust any online service to be permanent in their behavior. I just enjoy it while it lasts.
> it's all of the other stuff that tends to get hosted together with the code (issue tracking, pull requests, releases, potentially even CI) that's a lot worse to have to migrate over. Plus, anywhere a link was published to the repo is going to point to something that no longer exists.
I think most of them provides API endpoint to access those data or some export tools. And as far as links go, I would prefer to have a webpage for any relevant projects, something like the one for msmtp [0], tied to my own domain.
[0]: https://marlam.de/msmtp/
1. This precedence is based on law or on the abscence of it.
2. Slippery slope fallacy.
3. Anarchy or Tyranny are not the only modes of operation. -- See point 2
Did you know GitHub removes things they deem to be malware? Malware is a very subjective classification, so who watches the watchmen? How do we know they won't start removing things that are deemed harmful just for competing with Microsoft products? Hell, letting them have a mechanism to moderate the site's content at all is a huge risk to society's freedom and continued existence. Aliens could arrive any time and shoot mind control beams at the GitHub admins which compel them to delete everything at once. The only way to prevent the future holocaust is if I'm allowed to typo squat GitHub repos to distribute viruses and someone else has to pay for it. If it's not safe for me to do that, GitHub is not safe for anyone.
Removing malware or illegal content is very different than removing things for political reasons.
What is and isn't illegal is the very definition of political
And also like... this is how literally any community works? You "invest your free time" in it, then you do something that goes against the values of that community... yeah, of course that big bad "shadow group" (in other words, the admins/moderators) is going to do something.
The point is that the stated values of said group changed.
The values didn't change, they just made them clear.
Yes, this happens with most communities. It's called growth.
As an example, HN now has this in the guidelines:
> Don't post generated text or AI-edited text. HN is for conversation between humans.
Does this signify the end of HN? Or does it represent understandable growth in an evolving world?
It's not analogous at all. I moved my project to codeberg because the values were one way, and GitHub didn't match my values. I pour my heart and soul and time into my project. I also used AI tooling in a fashion as I saw fit. Now I'm reading I'm not welcome -- retroactively -- because I touched "bad" tools.
So I had to spend the last 3 days dealing with all the fall out from that and moving my project -- which was not insignificant -- docker images, debian package hosting, documentation references, existing in-flight pull requests, blog posts, build/release processes, my CI configuration, etc. etc. I'm nowhere near done.
An individual making an AI slop comment on HN loses that one comment to moderation and moves on.
I'm with you on this. I started moving projects to Codeberg when my understanding of the rules was that it was friendly to FOSS projects. Now it's explicitly unfriendly to a set of FOSS projects (albeit of a kind that I didn't participate in anyway) and users who use a particular kind of tool while working on their otherwise OK FOSS projects.
Now you can have a perfectly legal GPL-licensed project that's not welcome because of the category it finds into or because of its devs' tooling. It's their site and their rules, sure, fine, but it does mean I'm not going to host my policy-compliant projects there. Are they going to get banned next month? Who knows.
To be blunt, I think you may have misunderstood the values to begin with. A motion being passed by a majority is a lagging indicator to the values of the membership, so if they changed they will have changed a long time ago.
The Codeberg blog has articles critical of AI published back in September 2024[0]:
I don't even use Codeberg, and I've understood it was at best wary of AI for years. This shouldn't have come as a shock to you.[0] https://blog.codeberg.org/letter-from-codeberg-software-is-a...
This is basically the argument I've seen over and over again from the Codeberg community. Boo hoo, you misunderstood the writing on the wall and now you're stuck doing a bunch of work. Goodbye.
I mean, more power to you (the community) in setting up your fiefdom, but by God do I suspect there's going to be some good people who leave, not wanting to participate in that sort of dialogue. I don't understand why it's so hard to show a little understanding and compassion.
The goal is to reduce the amount of use of the platform, and I think imposing this rule is an excellent way to do that. Not sure I'd call it genius, though. There's a fair likelihood of unintended consequences.
these things always end with community disintegration. Eventually, some high profile user will cross the undefined line or leaders change and want to rewrite the rule then half the community will side with them and half will not.
Yes, every community disintegrates. Eventually. Individuals also disintegrate. Over the long run, planets and stars disintegrate. Best not to think about it too much lest you go insane.
I think we should point out that Codeberg is a e.V. and that they have a Mission Statement [1]
> The mission of Codeberg e.V. is to build and maintain a free collaboration platform for creating, archiving, and preserving code and to document its development process.
[1] https://docs.codeberg.org/getting-started/what-is-codeberg/#...
As I understand it, they do not ban Crypto. Crypto is just an example for a category of projects that harm there reputation. The formulation for AI generated code is different. It does not point out AI / LLMs as example for copyright issues. It does it the other way around.
Seems like a recipe for endless division and meta arguments rather than doing anything useful.
Reminds me of internal political disagreements.
This is a lesson I learned in life: When you want to build something social (in this case a community, but it could also be a romantic relationship, a friendship...), the best thing you can do is to polarize as quickly as possible.
Too many people are afraid to offend and waste everyone's time going along with things they don't really agree with and end up pretending to be somebody they're not to get along with people they don't really like.
No! Get your most polarizing takes out quickly. Either you and the other person stop wasting time because there's no fit, or you will deepen the relationship and both know it's a good fit. Either way, it's better for everyone involved.
There's enough people out in this big world that you don't need to always make everyone happy. It's much better to quickly comb through to find the gems.
> There's no issue with this. This isn't the supreme court. Your income is not on the line. It's a code hosting repository. Go to GitHub if it bothers you.
Codeberg is best known as a place for people to go when they're fed up with GitHub. People who are fine with GitHub's various issues wouldn't be looking at Codeberg in the first place, as far as I can tell.
There is GitLab, though. And Forgejo and a bunch of others.
The big thing I read is Codeberg is for community projects explicitly now, not just your random non-commercial code.
Personal use cases are now “might go unnoticed” and therefore not encouraged. Jeff’s dotfiles repo will probably never have a community, so it’s technically against the rules for some future cleanup initiative.
Most of my repos are random personal projects that are not supporting any larger open source thing, so I couldn’t look at (or recommend Codeberg) as a GitHub alternative for anything but an established project.
I think that’s all an acceptable choice, but it’s not being discussed much.
> The fact that the rule is open for interpretation means that they can effectively shut down malicious actors
This also mean that some kind of charismatic leader can become a de-facto tyrant.
And btw, lone leader on top, bending the rules at will… that’s a trait all dictatorships have in common (hitler, mussolini, stalin, mao, etc). Just saying.
Ans that's why I don't use Python or Linux.
This is the problem with choosing platform for moral — instead of practical — reasons.
Platforms that are known for making moral judgments will likely keep making moral judgments in the future. Eventually, they'll make a judgment that you heavily disagree with.
This is why communities founded on ideology tend to shatter, while those founded on a shared goal tend to survive. Linus doesn't care whether you're a GPU expert at AMD, a 16-year-old in Nairobi, or a nazi with a swastika tattooed on your arm. He cares about the quality of your kernel patches.
He does have some choice words for you if you're Russian, though:
> As to sending me a revert patch — please use whatever mush you call brains. I'm Finnish. Did you think I'd be ‘supporting’ Russian aggression?
(from https://therecord.media/linus-torvalds-russian-linux-kernel-...)
I think this shows that there is a blurry line between moral and practical judgment, and that the type of judgment might not even matter in the way you think. Sure, this episode may have been started by OFAC sanctioning various Russian companies, but Linus is happy to invoke more personal moral terms; and ultimately whether it's practical or moral doesn't matter, a bunch of Russian folks found themselves without a commit bit seemingly at random and the same could happen to anyone regardless of a particular platform's or company's judgment.
his personal stance did not lead to anybody being excluded though.
yes it did? He banned a bunch of developers with russian email addresses.
You are saying this like pursuing profit above anything else is not a moral judgment one can heavily disagree with.
No; they aren’t making a judgement. They’re just describing a relationship between the nature of a platform and its likely future behaviour.
Who is “they“ is your sentence, I’m not following, sorry
The person you were disagreeing with.
Ah but I never said they made a judgment?
Pursuing profit is not a moral judgement.
It's at best neutral, and at worst considered to be immoral.
Nobody says "I need to make a billion dollars because of my morals".
It is, especially the “above everything wise” definitely is an ethical decision, and one many often disagree with.
Is Linux really about "profit before all else?" That seems like a moral judgement that many would disagree with.
I didn’t say Linux is about this? I’m not following you.
Is the fact that churches are some of the longest surviving communities accounted for?
(and of course Linus cares, like any unaccountable asshole he can be triggered by a way more trivial thing than a swastika (see the side comment for an example)
Perfectly summarised. Also recently saw so many people moving from GitHub to Codeberg because of anti-US and sovereign European tech movement and I wonder now which platform they are going to move again
They need to self host if they want sovereignty. This thing about relying on an external service and demanding total privacy and control is irrational.
I think this would be a good time for another European alternative to present itself. Outside the US, but more pragmatic than Codeberg.
It will exist if you create it
I might. But I'm currently looking at CodeFloe. It sounds a lot like what my best attempt might end up as.
Deciding that you "won't make moral judgments" is also a moral judgment. It's just a less obvious one. All platforms have made moral judgments in the past, and will continue to do so in the future. Some of them just don't realize it.
Its a choice they're totally allowed to make, I just wish we'd get an updated mission statement[1]:
"The mission of Codeberg e.V. is to build and maintain a free collaboration platform for creating, archiving, and preserving code and to document its development process."
Amended to include at the end ", as long as you use the tools we find agreeable."
[1]: https://docs.codeberg.org/getting-started/what-is-codeberg/
I have no particular problem with this decision/policy on the part of Codeberg.
I do have an issue with their "tainted code" project that labels various FLOSS projects as "AI tainted".
ardour.org has the follow AI policy statement:
---------
Using LLMs to generate code for Ardour
A word about licensing, copyright and LLMs ("AI"). We require all code contributed to Ardour to be licensed under the GPL (you may retain copyright, but the code must use the GPL). Currently, following Thaler vs. Perlmutter in March 2025, no LLM-generated code may be copyrighted in the USA. The GPL relies on copyright, and cannot function if copyright is not held. Consequently, if you use an LLM (Claude, Codex, CoPilot etc. etc) to create code, it cannot be included in Ardour.
If you want to use LLMs to explore and understand some of Ardour's source code, feel free. But we cannot accept PRs or patches or contributions that are not licensed under the GPL, and if you choose to be honest about it, this means that no LLM-generated code can be accepted by us.
---------
Yet this codeberg-adjacent project considers this project to be AI-tainted. If you're going to do something, do it right or don't do it at all.
They didn't consult a lawyer when they wrote that. The idea that public domain code that is not copyrightable is incompatible with the GPL is simply not true. sqlite is in the public domain and not copyrightable, for example.
Their decision leans too heavily on what that court decision said. It was actually fairly nuanced.
The court explicitly declined to address whether Thaler could be considered the author because he created the AI, because Thaler originally listed the machine as the sole author in his registration application.
It is clear that works generated entirely by AI without meaningful human contribution are not eligible for copyright protection, though it does not resolve questions regarding AI-assisted works where human input is significant, which is a predominant use case with AI-assisted coding.
Finally, it appears they're unfamiliar with copyright rules around non-copyrightable works that are combined with other creative copyrightable works. The result is almost universally copyrightable. I'm hedging because I haven't thought through the cases exhaustively, but I can't think of a single case where the resulting work would not be copyrightable.
So they have modified their policy because they do not understand either copyright law in general or Thaler vs. Perlmutter in particular.
I think we will be seeing a lot more of this. There were plenty of examples of developers failing to understand copyright law and the implications of licenses before LLMs and that was with well established law. With some genuine uncertainty around LLMs, more variation between jurisdictions and some criteria that are less clear its going to get a lot worse.
> sqlite is in the public domain and not copyrightable, for example.
And therefore is not licensable under the GPL; because no license is needed at all. Ardour's goal isn't to be "compatible with" the GPL (in the sense of allowing binaries to be built containing both GPL-licensed code and their own code); their goal is to be covered by the GPL license for the purpose of constraining what actions users can take. This requires the code to be copyrighted, because copyright law is the mechanism that allows them to require users to agree to a license.
You're ignoring that adding non-copyrightable code to a collection of copyrightable code makes the collection copyrightable. The folks that wrote the policy are also ignoring that. It's simply a legal misunderstanding, because (I suspect) they didn't actually consult a lawyer, and their reading of precedent is overly broad.
Ardour does not seem to require a CLA or copyright assignment, so the maintainers don't own the copyright over contributions either way. The contributor would not be able to sue someone over a GPL violation if their contribution is public domain, but what random small contributor is going to participate in a GPL infringement suit anyway? Doesn't change the maintainer's rights to do so with any of their own human-authored contributions.
I don’t even mind forges taking a stance on AI. That’s fine. But I agree with Armin Ronacher here: banning projects that “mostly consist of code written by generative AI tools” is like the infamous Supreme Court “I know it when I see it” line. Yes of course I know a project mostly written by generative AI when I see it.
Exactly.
> The line is open to interpretation precisely where it needs to be enforceable.
It is way too arbitrary and open to interpretation and biases.
“For my friends, everything; for my enemies, the law.”
I don't understand why people are referring to Codeberg as competitors, or going to competition. They're WANTING you to go back => GitHub and consume the GitHub's resource's if you do not conform to their traditionalist nonprofit view.
They are not trying to retain you as a customer or even have customers ! They do not advertise ...! I am a [relatively responsible] utilisateur of AI and I truly do not understand the entitlement of some people.
It is a nonprofit, the vast majority of people complaining are not voters, donators, or members, while the vast majority of nonprofit members have chosen this view.
It's the same thing as the "jqwik" issue.
My personal opinion, is that the people clutching pearls are _scared_. There is a nagging feeling that deep down, they know the machine gods they love so dearly are _wrong_.
That if actions like these pick up momentum, if their "experiments" work, and more and more people start realizing that alternatives exist, the current places will end up as mostly "dead malls" as the people have moved on to the new places. And they will be facing at a gate where the people guarding it know where they came from, and why they are now asking for entry.
That's a nice xenophobic fantasy. There's a reason why border guards are known for their power trips.
I think introducing unpredictable new prohibitions is a much bigger problem than the specific bans on cryptocurrency or AI written code. Especially when it leads to existing code becoming unavailable and doesn't just ban new contributions violating the rules.
They weren't exactly unpredictable. People didn't predict it because they're used to dealing with faceless machine platforms instead of human platforms. GitHub bans projects rarely, and when it does, it's without any warning or rule change.
Related:
https://news.ycombinator.com/item?id=49015635 Protecting our FLOSS commons from LLMs (blog.codeberg.org) 1 day ago, 144 comments
https://news.ycombinator.com/item?id=49003386 Codeberg bans vibe coded projects (codeberg.org/codeberg) 2 days ago, 279 comments
Thanks! Macroexpanded:
Ah, the Codeberg Drama - https://news.ycombinator.com/item?id=49034928 - July 2026 (51 comments)
I regret migrating to Codeberg - https://news.ycombinator.com/item?id=49021856 - July 2026 (474 comments)
Protecting our FLOSS commons from LLMs - https://news.ycombinator.com/item?id=49015635 - July 2026 (144 comments)
Codeberg Bans Cryptocurrency Projects - https://news.ycombinator.com/item?id=49015588 - July 2026 (638 comments)
Codeberg: ToU extension to prohibit LLM-extrusions - https://news.ycombinator.com/item?id=49003465 - July 2026 (77 comments)
Codeberg bans vibe coded projects - https://news.ycombinator.com/item?id=49003386 - July 2026 (279 comments)
Codeberg should host code, and should not be bothered by which tool the code is written. Let's wait until GitHub starts accepting only code written in VS Code or Visual Studio because those are Microsofts tools.
Want to use other tools? Move somewhere else.
Codeberg going down the path of Blueskyification
A notable surprise with Bluesky over the past few months is that the anti-AI sentiment has decreased slightly as it becomes harder to argue against impactful use cases.
At the start bluesky was actually pretty nice, but now it's just this weird bubble where disagreements don't exist and when I checked out like a month ago there was just so much damn p***.
Aren't there good filters for that? It's one of their core differentiating features - the ability to use custom content filters. Or did you check with no protection enabled?
I genuinely have no idea what you mean or what talking point this might be referring to.
Bluesky, from my daily use, seems like it’s thriving and full of varied and interesting people.
It is a bad rule. Their community and their choice but I hope we get some fresh options in the open source universe.
This sounds great. We've hit some kind of scaling problem where "algorithms" are not up to the task of finding worthwhile stuff among the solar masses of junk being pumped out daily. And even if they were, the amount of exploitation is so fathomless that I don't think it could ever work. It's time for society to rediscover the power and immense value of real curation. Even with non experts, just looking through another human's playlist of recommended songs beats the crap out of algorithms. I've found so much more music that way than with youtube suggestions or whatever spotify does.
This decision may also be related to infrastructure. Even if Codeberg wasn't against heavy LLM use for moral reasons, their limited infrastructure would have a hard time keeping up with the massive output that LLMs are capable of.
For once I wish they would've taken a more technological solution to this social problem and enforced more limited CI time and disk space quotas by default. Seems like it'd be fairer and require less active policing. My read from their post was that the resolution had more of an ideological bent than being just about resources, which is certainly their community's right to have.
I am a heavy viber that uses a lot of resources in github now after LLMs compared with pre-LLMs for my personal projects.
I find great that there are places that are still human first and sensible about community and free software instead of just providing technical infra. Gitlab and Gitea are already open Github alternatives: open source core projects turned open sourcish business with free as in beer plans that get monetized with user data, loss leadering, business offerings and other sassy business models. Codeberg is deciding to not be that. Other projects can decide different and its users should evaluate the tradeoffs and choose.
> I could not reliably assign authorship percentages to many of my own recent projects.
This feels like a strange statement. Unless you are going out of you way to disguise AI contributions it's extremely simple to do some git magic to get an exact percentage. But you can also just ask AI and it'll tell you directly. And even if you've tried to disguise it I'm fairly certain that if you set aside 10$ for tokens, it'll be able to deduce how much is AI even with someone having tried to disguise it.
codefloe[1] also runs forgejo.
[1]: https://codefloe.com/
My only complaint about their site is that it mentions a generous free tier, but I have no way to seeing what exactly that free tier is.
Edit: Found it, but it wasn't obvious the navigation path to it. Would be nice if "free tier" on their home page was a link where it was there and clear.
> It can choose to be a smaller community with a stronger political identity, but that is a different ambition from being a broad and dependable European alternative to GitHub.
Wrong category. It doesn't want to. It's not a company. Might as well act on their beliefs.
Devs are so spoiled with free CI, running nonsensical workflows all the time for free. Imagine that multiplied by magnitudes thanks to AI. Just a waste of resources.
Devs are so spoiled. They first tried to sell themselves as morally better than GitHub, and now the spoiled devs ban who uses AI..
Spoiled with freedom of association and freedom of speech
The facts are missing in the Article. The changes really made by Codeberg e.V. are:
First:
> 7. You must not share projects that mostly consist of code written by "generative AI"-tools (including services such as Claude, OpenAI Codex). Such projects having an unclear copyright status (see requirements § 2 (1) 1 and § 2 (1) 3) and furthermore have little safeguards to ensure that they do not include harmful code (c.f. § 2 (1) 5).
Second:
> - Content that harms the reputation of Codeberg, such as cryptocurrency related projects.
> Since I want GitHub to face competition I have thoughts.
Github allows private repos, organization access control etc. I don't see how Codeberg is competition.
Codeberg was competition for the EU, in the topic of tech-sovereignty which is a incredibly hot topic in the EU. With the US abusing it's position not only recently but years past, the EU is accelerating funding and plans to decouple from anything the US can mess with legally. Much of that sentiment has been spreading all over the EU governments and software communities.
GitHub for all its purposes remains a platform hostile to EU interests long term.
I don't think they were saying that there aren't reasons to not use github, in or out of the USA, but rather that codeberg was already not a contender for many use cases due to their existing rules restricting private/non-OSS repos.
It's apples to oranges. No matter how many reasons one has to avoid github, Codeberg is only an alternative for open source hosting.
Are there any alternatives that you could use for closed source development that aren't owned by US entities?
Self hosted Forgejo
> I wish Codeberg were more forward-looking here: willing to host the Open Source software of tomorrow, not only software made in the ways its community approves of today.
Is "vibe-coded software" really going to be the "Open Source software of tomorrow"? The opinion stated in the blog post announcing this change is the opposite:
"We believe that LLMs endanger the free/libre software ecosystem as a whole." and this is expanded in more detail later in that blog post (starting at the "Collaboration at danger" section).
This reeks of the same FOMO strategy seen often in cryptocurrency projects (which, unsurprisingly, have also been banned from Codeberg at the same time): stating that their way is going to be the future, and that those who don't follow it will be left behind ("have fun staying poor").
> But these tools are also becoming part of how software is made. The Open Source world needs to figure out how to engage with that future, not just divide into camps.
No. They don't. This isn't a "you just have to deal". You can say no. It's OK to decide that it's not worth the smaller amount of positives to deal with the massive amount of negatives. If you disagree, start your own community.
There’s a lot of hue and cry over what’s just Yet Another Bulletin Board because it has a git repo attached? Just host this stuff yourself guys. It’s trivial these days.
I have my own remotes with my own CI and it’s all ultrafast because I can just make assumptions about the way I write software.
None of these platforms are necessary for development. Just take from them what you want and build your own universe. This grade of CRUD software is generally solved.
There are places you need other people. Stuff like ffmpeg or GPU kernels. But other kinds of code is kind of unimportant. Not being part of this community loses you nothing.
Is there any evidence of division? Are all the vibe-coders going to some codeberg2? One of the main complaints is that the banned projects don't have human users proportionate to their activity, so if they all go to codeberg2, then that forge would just be full of projects nobody cares about.
Are there any projects on codeberg that are getting banned and actually have significant human interest?
If codeberg decided to kick out spammers, nobody would decry the loss, so what's being lost here? Is it strictly slippery slope?
Nobody has been banned yet, AFAIK. So nobody can answer your second paragraph question. But plenty are choosing to leave.
The judgement you're talking about here ("banned projects don't have human users propostionate...") is going to require evidence? At this point the software community is absolutely full of these tools and people who use them.
If the specific thing banned was projects like that, fine. But that's not what the clause added in the PR said.
> At this point the software community is absolutely full of these tools
That is self-evident.
> and people who use them.
This isn’t.
Maybe it will push more people and more effort into the true decentralization since the Forgejo/Codeberg federation[1] effort is way too slowly moving. Radicle[2] it is.
[1] https://codeberg.org/forgejo-contrib/federation/src/branch/m...
[2] https://radicle.dev/
Just self host if you want freedom. I switched to Codeberg to maintain a semblance of project discoverability while having freedom from Microsoft but I don’t need the code police telling me what I can build or how I can build it. My open source values are more WTFPL "freedom to do whatever you want" and Codeberg is more AGPL "freedom to do whatever we allow you to", which is totally fine.
It seems to me, Codeberg sees no value in AI slop and doesn't want to spend significant amounts of its limited resources hosting it.
The author talks a lot about what they want from Codeberg. And I suppose it's fine to wish someone (on the right side of the ocean) would provide services, tailored to your personal desires, for free. But don't be surprised when no one takes you up on it.
For myself, I don't see a point to Codeberg simply being a clone of GitHub. GitHub is free, which is great, but it's also pretty shitty. Why do we need more of that?
I see. We should not divide into camps so the Austrians can sell their AI startups.
The only people dividing and ruining open source are those paid for it by AI companies. Sometimes indirectly, like NumPy through its various sponsors. Then suddenly a bureaucrat appears on a mailing list or elsewhere and announces that AI contributions need to be allowed.
Fully agree. I am against AI slop, LLM overhyping, and cryto scams, as much as anybody, however:
> the main property I need from infrastructure [is for] it to be predictable, dependable, and reasonably neutral towards the legal Open Source software hosted on it.
I do not like to build on a foundation that could be ripped away from under my feet at any moment for arbitrary reasons.
> I do not like to build on a foundation that could be ripped away from under my feet at any moment for arbitrary reasons.
Do you only build and publish software in a vacuum running on hardware you own? There is no online community in the western hemisphere that isn't subject to any kind of arbitrariness, be it internal (community, like here) or external (legal or infrastructurally, like on github).
Yes, but the world is not black and white. It’s all shades of gray. Codeberg just darkened their shade quite a bit.
I reiterate what I said in another thread (and invoke Don Mattrick): If you want AI slop, there's a perfectly fine*** product for you already out there - it's called GitHub.
For those of us who are tired of AI messing with things, stop complaining and let this be. Codeberg made a great decision for those who are just tired and over AI coding unnecessarily taking over everything.
> The Open Source world needs to figure out how to engage with that future
How ironic. Why is the burden of labor on the open source community? They have to do more volunteer work for your AI satisfaction, even after training the very models you depend on? I think proponents of AI are the ones who have the burden of figuring out how to engage with the open source community. Use your own tokens to figure it out. Surely AI is up to the task. Your frontier models have the entirety of human diplomacy at its fingertips to steer you towards the path of success. I wish you luck.
I'd rather have a way to quickly ID what deltas were generated and which were not but that will never happen if it becomes a weapon to ban a project.
> More importantly, LLMs if done and used well, should be welcome to all of us. They could be used to reclaim control and power, away from large corporations and institutions.
Is it just me, or does this read rather naive?
This decision is consistent with the community Codeberg has attracted of late. It's smaller, independent and takes stances that are not in line with competitors like GitHub. If you want something less reliable, at larger scale and filled with AI projects that GitHub is clearly encouraging, you should use GitHub.
Genuine question to AI users upset by this: if you trust AI to generate your code, why don't you trust it to help you "deploy https://forgejo.org/ on cheap VPS make no mistakes"?
Related:
Codeberg bans vibe coded projects
https://news.ycombinator.com/item?id=49003386
Codeberg Bans Cryptocurrency Projects
https://news.ycombinator.com/item?id=49015588
Protecting our FLOSS commons from LLMs
https://news.ycombinator.com/item?id=49015635
I regret migrating to Codeberg
https://news.ycombinator.com/item?id=49021856
So, only self hosting. Gonna mirror my github repos on my synology!
You should be doing that anyway, no external hoster can be a replacement for having your own backup.
So, i will self host all of mh gh mirrors on synology. Never liked strange us/eu "communities" with their strange values anyways
So are they going to have "AI" police this? Heuristics about check-in size and scope are still AI, albeit "classical AI."
The low hanging fruit are claude.md, commits co-authored by Copilot, etc.
on April 1 next year don't be surprised when lots of repositories come out with "surprise! i was AI all along". How are they ever going to know? Seems like a prompt like "remove all traces from this code that it was generated using AI" would one shot the whole repo and make it virtually impossible to justifiably claim it was vibecoded.
Considering the fact that AI-written code is the future, Codeberg has damned itself to be a niche project forever.
That's the beauty of the internet, everyone can find their space. Even the buggy whip makers.
All these issue and gitlab is still there, fully working and fully functional, open source and ready to host your stuff without useless drama.
It’s also got a great ci/ci experience.
I write almost all code with GenAI nowadays. I'd like to say it's not slop, but I guess now there's no way to get Codeberg to agree.
"Write" is not the correct verb for how you produce software if you get it from GenAI.
May I suggest a few replacements? "Plagiarize" or "excrete" are both more fitting in different ways.
> There are serious questions about copyright, labor, energy use, slop, and maintainers drowning in generated contributions. But these tools are also becoming part of how software is made.
This first sentence is such a weasel-acknowledgement that is used to just effectively ignore all of it.
Agree substantially here, as I said elsewhere: the honest and respectful answer from the Codeberg community would have been: add resource limits, quotas, and a reviewed application form for new projects.
And maybe a statement that: Pre-existing clearly "vibe coded" projects providing little value to the community may be subject to review. Our resources are limited. Don't abuse them.
Instead it got framed in moralistic terms in many places, the actual voting process actively resisted all calls for clarification ("if you're a vibe coder you'll know, so stop trying to ask for clarification", the tone in online forums became nasty, and the result is both ambiguous and culturally poisoned and I decided to move my project ("good riddance!" call out the victors, "we got one of them!")
I do think the one substantial argument that was made through the threads, was around IP ownership laws. However I remain unconvinced.
Frankly if there's anywhere where I would most like to see AI generated work -- which is frankly based on weights generated from the broad and indiscriminate harvesting of other people's works -- is in copylefted software that takes the distillations of the efforts of the community and makes it available as free software with free software protections.
I understand not everyone shares that opinion, and they're welcome to different interpretations.
Like the author of the article I'm saddened to see this split. I am honestly curious to see where we are in 2-3 years on this topic.
> Frankly if there's anywhere where I would most like to see AI generated work -- which is frankly based on weights generated from the broad and indiscriminate harvesting of other people's works -- is in copylefted software that takes the distillations of the efforts of the community and makes it available as free software with free software protections.
I agree wholeheartedly with this. We have large corporations stealing work left right and center. Lets utilise that to our benefit and make open and free versions of all software. The crux is that to do that each individual has to pay currency to use this tool. Maybe there is a place for collectives to collect money and to utilise llms to free software. (What strange times we live in.)
A slight aside, open source lived and breathed the idea of ownership of software, be that a council, a handful of committed maintainers, a BDFL and all things in between and adjacent. There isn't really much wrong with "vibe coded" software, the issue seems to be owning that software. Things feels fundamentally more fluid, based on specs and more importantly tests that actually define the features that the software is expected to perform. Not ownership, but more like a contract.
Resistance to all forms of oppression is a duty.
A great way to ensure that most developers will not be using Codeberg going forward. While there's a ton of drama around LLM use right now, anybody who's actually used this tech in anger understands that these tools really do work, and that they will only be used more in the future. I guess Codeberg is angling to become a small artisanal community, and there might certainly be a niche for that.
If the outcome is that most vibe-coders leave Codeberg alone, that seems like a win for Codeberg.
The outcome is obviously going to be that most coders leave Codeberg because LLMs are not going anywhere, and they will continue to be integrating into people's workflows. But perhaps Codeberg intends on becoming something akin to a digital Amish community, and there's nothing wrong with that.
"Not going anywhere" is such a strange argument. Tomato juice isn't going anywhere, does it mean I'm going to be guzzling it at some point in my life? What?
If you've made your choice about LLMs based on morals or ethics rather than the quality of the tooling, you're not going to change that choice because the models haven't disappeared or have improved, unless you're forced to (but that forced bit applies only to commercial open source which really just isn't on Codeberg at all in the first place).
It has been long enough that just about everyone has made that choice by now. Those who are okay with this won't magically flip and start using LLMs and grow a reason to leave.
Every moral or ethical argument ultimately boils down to problems inherent in how any piece of automation is used under capitalism, and has little to do with technology itself. If you spend a bit of time thinking about morals or ethics you'll quickly realize that LLMs are not the problem.
And no, a year is absolutely not long enough for everyone to have made a choice about this tooling. Claiming any sorts of absolutes about completely new technology that's rapidly evolving is absolutely delusional.
The current adoption of LLMs by developers seems to be a bit more than 20%.
https://www.pewresearch.org/short-reads/2025/10/06/about-1-i...
This is just for professional jobs (and is 9 months old) of course.
That's an incredible rate of adoption given that this stuff has only been really usable for the past year or so.
It sure is! But I guess it would be less fun to say "1 out of 5 developers will not be using Codeberg going forward". Especially since that may well be an intended effect of Codeberg's decision here.
As long as AI is a source of environmental damage, cognitive harm, and a cause of rising hardware costs, there will always be a sizeable set of developers who want nothing to do with it. The headline should be "AI Divides".
There's no basis for thinking that LLM use will not continue to grow going forward. The fact that 1 in 5 developers already use this technology only a year or so after it's become really useful shows that adoption is growing rapidly.
Meanwhile, the arguments regarding environmental damage and cognitive harm are rather spurious. You can already run models like Qwen 3.6 27b on your laptop today, and it can do serious work. And the efficiency of the models only keeps improving. Meanwhile, studies showing cognitive effects you allude to have used laughably small samples, and these same types of arguments were made every time the way people work with code changed. I'm old enough to remember people screeching how you weren't a real programmer if you didn't use assembly, or that GC was ruining programming.
There's always a group of people who made a certain set of skills their whole identity who get really upset when those skill become automated. There will always be a sizeable set of current generation of developers who will want nothing to do with it, but people getting into development today will treat LLMs as a part of their regular toolkit. At the end of the day it's just a tool like any other.
> There are serious questions about copyright, labor, energy use, slop, and maintainers drowning in generated contributions. But these tools are also becoming part of how software is made. The Open Source world needs to figure out how to engage with that future, not just divide into camps.
Everyone who says stuff like this is just burying their head in the sand. That is like saying that there are serious questions about the ethics of buying sweatshop-made clothes, or driving a gas-powered car, or robbing grocery stores --- but those are becoming part of life, and so we need to "figure out how to engage with them".
That doesn't mean you have to upend your life to fight the good fight. You can say "I want a world without X, but right now I need X to live, so I do X, while maintaining an awareness of it and looking for opportunities to do less X". You can say "I want a world with some kinds of X but not others, so I do the X that I'm okay with". You can even say "I think X is ethically fine so I'll do it". But it just doesn't make sense to say "there are ethical issues with X, but those only affect how we engage with it, and have no bearing on whether we think people should need to engage with it at all".
There can't be ethical consideration of "how" without at least considering the ethics of "whether".
> There are serious questions about copyright, labor, energy use, slop, and maintainers drowning in generated contributions. But
Way to not at all engage with any of that and just go on your merry way...
Sure, if you fail to acknowledge that the other side has valid points, it is easy to act like you're completely right and they are morons, and we all need to Kumbaya together, and forget about their views.
The way I read the article was not that the author disagrees with the concerns you mention, but instead that they believe Codeberg's recent actions fail to address these underlying concerns while adding uncertainty for users of their platform (which they fear will ultimately hurt codeberg). To me, it reads as a call for further discussion and clarification rather than a dismissal.