It’s completely a coincidence that all games are no longer working in Lutris here, on multiple machines, after upgrading from 0.5.19 to 0.5.20. Weird.
I downgraded and everything works again. I did not try 0.5.22 or the quickly removed 0.5.21.
Oh man, you are right. I went fron 5.18 to 5.20 and nothing worked anymore. I spent hours troubleshooting before I reinstalled the current game I was playing. It worked but it runs noticeably slower. For a newbie, how does one downgrade? Assuming there is a command or do I have to uninstall first?
Someone suggested the program Warehouse to me, but I haven’t tried it. On Arch, I still had the version I wanted in my package manager’s cache so it was a single command.
sudo pacman -U file:///var/cache/pacman/pkg/lutris-0.5.19-9-any.pkg.tar.zstIf you are using the flatpak (Bazzite, Steam Deck, etc.) unfortunately, it’s more complicated.
- Exit to desktop mode.
- Open Konsole (it’s in System in the main menu).
- If you haven’t set a root password yet, run
passwd, make it reasonably secure and don’t forget it. I believe setting a root password enables the Deck to be controlled remotely over ssh with said password. Be safe. - Run
flatpak remote-info --log flathub net.lutris.Lutris. Lutris was installed as system for me. I think that is the default, so probably choose 1 for system if it asks. - You will see a commit with the subject “Update Lutris to 0.5.20”. The previous version is in the list right after that. Note that hash of 64 hexadecimals.
- Run
sudo flatpak update --commit=19ee79d455b8e50f057911a2bba279efcb960ee6d565f794e9c9d41c290dcd14 net.lutris.Lutris, supply the root password, and accept the changes. (Use the hash from step 5.) - Run
sudo flatpak mask net.lutris.Lutrisand supply root password to prevent Lutris from being updated. We will probably have problems in the future when the flatpak environment gets deprecated,sudo flatpak mask --remove net.lutris.Lutriswould allow it to update again.
Wow, thank you for this! I really appreciate the detailed instructions!
After reading your comment, I tried it for myself, running “Age Of Wonders 4” through Lutris 0.5.22. Nothing happened. As in, literally nothing, game didn’t launch, and no error. Then downgraded Lutris to 0.5.19, and first I got a message saying that wine needed to install something, and then I got an error message saying “A java script error occurred in the main process”.
So the results of my experiment are inconclusive. I consider an error message a better than result than nothing visibly happening, because an error message at least tells me nothing its not working, instead of letting me wait and wait.
So, yes, it appears that the quality of Lutris has declined after the developer started using Claude Code. However, my experiment was just a quick and dirty experiment, and ultimately further research is necessary.
I propose the following experiment, keep in mind that this is basically a rough sketch of the procedure:
- Set up two virtual machines running linux, a and b. (TODO: Decide on distro)
- Install Lutris 0.5.22 on a, and 0.5.19 on b.
- Try out several games on both a and b, both installed and launched through lutris, and record how well they run.
lutris -dwill run it and print debug messages to the terminal.I think the root of the problem is that updating changes what WINE and Proton versions are being used, even for games that are already installed. That pretty much negates what most people are using Lutris for. (WINE prefix management)
I guess we know where to fork from.
I guess we know where to fork from.
Honestly: Why? Lutris Gnome headerbar UI sucks anyway. Looks and behaves like crap especially under Gamescope but in non-Gnome desktops it’s not too great as well. GloriousEggroll and team created umu launcher to make creation of that sort of graphical front ends much easier and a bunch of those popped up already. Might just as well migrate to one of those than to maintain yet another software fork.
deleted by creator
Just assume everything is AI generated and feel free to ignore the GPLv3 because generated code doesn’t have any copyright. See how he reacts.
That’s not how this works.
It is now, strycore made it happen.
The legal effect of AI generated code on software licenses is untested in court and AFAIK has no explicit laws. So really no one knows how it will work yet.
Just assume everything is AI generated
This is the part that will definitely not work.
If that AI slopper freaks out about alleging conplete lack of threshold of originality, it’s already a win.
The US Copyright Office has updated its guidelines:
If AI content is present, the Office will only register the work if the human contributions are sufficiently creative and if the AI-generated portions are supplementary or used as a tool under human direction. Essentially, they ask: “Is the work basically one of human authorship, with the computer merely assisting?” If yes, it can be protected (with a disclaimer that some content isn’t human-made). If no, if the AI’s role overshadows the human’s, then the work, or at least the AI-created portion, is not eligible for copyright.
In Canada, where I live:
So, can you claim copyright in an AI-generated work in Canada? As of 2025, the safest answer is: only if a human author contributed substantial creative effort to the final work. There needs to be some human “skill and judgment” or creative spark for a work to be protected.
If the AI was just a tool in your hands, for instance, you used AI to enhance or assemble content that you guided then your contributions are protected and you are the author of the overall work. But if an AI truly created the material with you providing little more than a prompt or idea, the law may treat that output as having no human author, and thus no copyright.
For now, anyone using AI in creative projects should keep documentation of their own input and creative choices. Emphasize the parts of the work where you exercised judgment or selected elements because those are likely what copyright will cover. And remember that copyright in AI-generated content is a fast-moving area.
https://www.foundationsoflaw.com/post/can-you-claim-copyright-in-ai-generated-works-in-canada
Makes sense to me.
The thing is, many of these guidelines are related to finalized products fully created by AI. As in, the AI produced a written or drawn work at the end of it that on it’s own is the product (Eg. an article or an image). This will probably apply to code in some reasonable way, but at the end of the day there’s only so many ways to write code since it’s syntax and not as flexible language. It actually has to produce something that works.
If you were to compare code written by two people at two companies, doing a very similar project, you wouldn’t be surprised to find two pieces of code doing almost the same thing in the same syntax, barring synthetic sugar like naming and coding conventions. Neither will likely have violated the other’s copyright since simultaneous invention is a thing. And if they happened to have similar prior experiences, it’s even more likely.
Likewise, the way the code was incorporated into a project as a whole might constitute a human contribution sufficient, and perhaps even the more important contribution. You likely wouldn’t retain the copyright on the specific snippet, but rarely are small code snippets enough on their own to claim copyright over to begin with. It’s the program or library as a whole that’s the finished product.
It looks like the issue submitter is trolling a number of projects on their personal anti-AI crusade. I would take it more seriously if they had reviewed any of the PRs and identified issues with them.
Yes AI slop is an issue (especially for maintainers) but it can still be a useful tool. If the maintainers want to use AI on their own code it should be their choice. Most forks fail because the righteous feeling of finally getting your own way on a repo you control usually falls away as you realise the people actually doing the work didn’t follow you.
Honestly the need for Lutris has gone way way down in the last couple years. I don’t know about forking it, but I think it’d be pretty easy to just avoid it. Less because there’s any concrete issues that I could point out, but more as a political statement and loss of confidence.
Here’s my issue with this specifically. It makes Lutris very vulnerable to being considered entirely public domain:
There is no settled legal status on the output of AI systems and it’s certainly something that does need clarification going forward. The law may treat asking an LLM to regurgitate it’s training data vs following instructions in a local context differently. Human engineers are allowed to use “retained knowledge” from their experiences even if they can’t bring their notebooks from previous careers. LLMs are just better at it.
As of March 2, it has been settled. AI generated works must have substantial human creative input in order to be copyrightable. Prompting the AI does not meet that requirement.
In other words, if the AI wrote the code, and you didn’t change it since then, it’s not yours at all. It’s public domain, no question.
US defaultism strikes again
I said, in the issue, I was talking about the US.
Prompting the AI alone does not meet that requirement. IE you can’t say “draw me a picture of a cat” and then copyright the picture of the cat claiming you made it.
You can say “help me draw this left ear over here, now make the right ear up here, a little taller, darken the edges a bit”, all with prompts, but with your sufficient creative input.
That’s not how the dev said he’s generating code. He said sometimes he does it without any intervention at all.
Also, that’s potentially copyrightable. That hasn’t been settled.
Your link doesn’t support what you’re saying in the slightest. Have whatever opinion you want, but don’t shovel up transparent bullshit to push your narrative.
TFA is about a a copyright on a work made by a purely autonomous device, and SCOTUS declining to hear a case doesn’t “settle” jack-shit.
Quoting further:
Thaler submitted an application to the US Copyright Office to register copyright in “A Recent Entrance to Paradise,” explicitly identifying the AI system as the author and stating the work was created without human intervention.
For now, businesses and creators using AI should continue to rely on the longstanding human authorship requirement. Under current law, works made solely by autonomous AI are not eligible for copyright protection in the United States. Ongoing cases also consider the amount of human input, including prompting or post-generation editing, required to register copyright in an AI-generated work.[12]
Companies should ensure a human contributes creatively and is named as the author in any copyright applications for AI-assisted works. To maximize protection, organizations should review their creative workflows and document human involvement in AI-assisted projects, particularly for commercial content. Organizations should continue to document the timing and scope of the use of AI in copyrightable works, for example by retaining prompts provided by the author. Internal policies should clarify attribution, ownership, the nature of creative input, and documentation requirements to avoid denied copyright applications.
Iteratively working on a codebase by guiding an LLM’s design choices and feeding it bug reports is fundamentally different from this case you’re citing.
If all you do is prompt the AI, “hey, fix bugs in this repo,” then you had no creative input into what it produces. So that kind of code would not be copyrightable, 100%. You can fight it in court, but the Supreme Court refusing to hear it means the lower court’s decision is settled law, and your chances of winning are essentially zero.
Whether code where you hold its hand and basically pair program with it is copyrightable hasn’t been settled. Considering the dev said he does it both ways, the point is rather moot, since for sure, he doesn’t own the copyright to at least some of that AI generated code.
OpenClaw is an autonomous system just like the one in that article, and the dev said that’s what he’s using at least some of the time. It generates and commits code without human intervention.
Glad it applies worldwide /s
Slop can’t be copyrighted, great. We don’t want slop.
“AI” has been known to present code from other projects and hence other licenses. It can’t become public domain unless all of that code was also public domain.
I’d imagine there have been more nonsensical (than AI = public domain) legal decisions that have had the full force of law for decades.
I recently dug around for a while, and if the copyright of works in the training data affects the copyright of outputs, no popular model can output anything that would even be close to acceptable for a contribution to an open-source project. Maybe if you trained a model exclusively on “The Stack” (NOT “The Pile”) and then included all the required attributions – but no ready-made model does that. All of the “open source” model frameworks that I could find included some amount of proprietary “pre-training” data that would also be an issue.
If AI output is NOT affected by the copyright of training data… there might not BE a (legal) person that can hold any copyrights over it, which is pretty close to public domain.
Good Sire, if we are talking about only the US, then that does not matter at all. Existing copyright law and established precedents (without involving AI) already covers this. The copyright of software is handled like that of literature, so the actual content is copyrighted. More specifically the sequence of words. In order to violate the copyright of a protected work, one just has to reproduce this sequence. It is not relevant, if it was reproduced by an AI, a human, God or your cat (:D). The only exclusion to this is fair use. Whether fair use applies must be considered by a case by case basis. There are four factors that are used in deciding whether it falls under fair use. And that is considering that portions of that code are not patented. If they are, then you are screwed no matter what (unless you are allowed to use that code).
Anyhow, you are opening yourself up for litigations for sure.
Now, is this a problem? Probably not. Copyright infringement is actually very very hard to spot, especially without automated tools (looking right at you, YouTube). Even if it is spotted, the owners of the copyright must use resources in order to enforce it. Considering that most of the code used in the training data is open-source, most of these owners won’t have these resources or at least aren’t using them (which is sad, because that also applies to the infringement of companies as well). You cannot lose, if no one sues. Whether you should risk it, is anyone’s decision to make.
For unprotected code… I guess, you are right. It could be one way or the other, but it does not really matter that much. At worst, people can use your code without adhering to your license. That would not mark the end of an project, the former definitely would.
Also on another note: Using copyrighted material in the training data of AI is considered fair use.
After their refusal to update their ssl keys for the Debian repo, I’m not at all suprised by this.
I’m now assuming it all is and deleting Lutris.
What a moron.
Oh yeah. Here’s another nugget:
Sometimes, I generate some code with Claude and commit by hand
Sometimes, I write code manually and ask Claude to commit
Sometimes, I ask OpenClaw to generate some code, which doesn’t put the Co-Authorship
Sometimes, the whole thing is AI generated from end to end
This is also a somewhat recent addition to Claude Code. I was kinda surprised when I first noticed it but didn’t think much of it, I was like “meh, I guess we’re doing that now, whatever, some people might take issue with it, whatever”. Also, do keep in mind that I love trolling people coming in my projects to complain about my methods.
For those who are anti-AI, it’s a safe assumption that any addition to the project has had some kind of AI interaction during the development process.
https://github.com/lutris/lutris/discussions/6530#discussioncomment-16088355
Sometimes, I ask OpenClaw to…
This person should not be trusted with anything.
So Trumps gonna give him the nuclear launch codes any minute now is what you are saying?
💯 this. I don’t mind using an LLM for certain tasks. We all do at the end of the day. However, OpenClaw is a different topic. This is just dangerous.
That is the real shame in all this. I’m certainly not updating lutris any more, because there is no way of knowing what you will install on your system.
You can trust humans (as in “trusting is an option”). You can never trust an LLM. And admitting that there might be unsupervised commits, being installed on possibly thousands of PCs is terrifying.
Glad I use Heroic instead. Time to check what their AI policy is.
Based on some PRs, they’re using github copilot to help with reviews but are generally against vibe coding
Now I’m really worried this software can wipe out my home directory
You’d think open source movement would take advantage of VC funded tools to fight against the big tech but instead we have literal ludites.
I have over 20 years of professional coding experience and I use Claude these days. Sure it makes mistakes and can write bad code but I’m not an idiot, I ran teams of dozens of engineers underneath me - I can handle a bot and fix it’s mistake. The maintainer of Lutris can probably too.
All I’m saying that this anti-ai mentality is fucking stupid and anyone who engages with it in such a binary way is fucking stupid too.
This isn’t just about anti-ai mentality, it’s the “I deleted the authorship so you can’t fork it out or prove that it’s causing issues”. This kind of insanity has been happening repeatedly on that project, it’s time to let it go and find new solutions.
This is clearly a response to the luddites? No?
It was a response to
All I’m saying that this anti-ai mentality is fucking stupid and anyone who engages with it in such a binary way is fucking stupid too.
First off, the luddites were right back in the day.
Second, just because you can use something effectively doesn’t make it good in general.
There are people who can have multiple credit cards for years and never carry a balance, or walk into a casino with $100, lose it all, and quit right there.
But most people can’t, and being one of the few that can doesn’t make it safe or good overall. Credit cards and casinos are still predatory and a detriment overall to the population.
I puffed a few cigs back in high school and college to see what all the fuss was about, didn’t get it. But I personally know multiple people that did the same thing, got hooked almost immediately, and took years to quit. Cigarettes are bad for you and highly addictive. The fact that they never hooked me doesn’t change that.
Third, I’m not sure how using LLMs is “fighting against big tech.” unless you just mean using their tools to build FOSS more effectively.
But that’s the whole point, it’s not at all clear that LLMs enable that for most people. In fact, there’s already quite a bit of data to indicate the opposite. That using LLMs results in worse code, worse development of skills like critical reasoning and problem solving, worse productivity, worse security, and undeniable environmental harm.
They were not right. What are you wearing now? You don’t destroy the machines - you take them.
How do you propose they should have taken them? Like, just lifted them in the night while the cops were busy jerking off?
By changing society and policy to benefit everyone rather than a few oligarchs? There are more of us than there are them.
In fact, this is exactly what we’ve been doing since the ludittes and you might even say it’s working rather well, slowly, but working nevertheless.
And this is something you have to do instead of breaking them? The policy pen must be held in both hands while writing, I take it.
And what you broke the machines then what? This teenage angst is cringe my dude.
Life continues as it was, or you build another. That’s kind of a silly question.
So FOSS Devs using ChatGPT are taking them…how?
Yeah, you can’t, because it’s not claimable, it’s proprietary software, even the “open” models are restricted in their licensing and usability.
Perhaps local models will make some positive difference in the future, but likely not. I suspect that decades down the line, we will see LLMs in a similar light as asbestos or Teflon. Things that were hailed as “miracle products” that seemed amazing at first, but had incredibly destructive side-effects that we only found out about long after the proliferation.
If something is overall destructive to society, I don’t want to “take” it, I want it gone.
I don’t want worker-owned casinos, or online gambling markets, or cigarette companies, I want those things gone. They don’t make us better off or more free, they harm and enslave us.
Good luck with that bruh
unless you just mean using their tools to build FOSS more effectively.
Until Anthropic eventually claims that they have ownership of any software written with the help of Claude.
but I’m not an idiot,
And you don’t think that makes you the exception among this cohort?
Not it’s an implication thay only someone lazy or an idiot would fail to fix mistakes produced by their bot assist. Worse case you just dont take the changes - you press the buttons.
You’re way too optimistic about the average person.
They are free to do what he wants on his repo.
We are free to fork if need arises.
Personally I don’t like projects not showing what AI has made. And most of Claude was made on stolen code. Its against the open source license they themselves use https://github.com/lutris/lutris/blob/master/LICENSE
But almost no one actually enforces the license until the big companies show up. I hope they change their minds, but until then, im going to stop using/contributing for a while.
Does anyone know which was the last version before the dev started shoveling slop in to the repo? The utter dipshit invalidated even the ability to license after that point, those releases are wholly worthless.
in 5 years from now there’s going to be totally coevolved but unique seed-lines for software. the once with AI, and the once without. how can you distinguish them? did the human that said it wrote them really write them? these problems aside, i suspect it will be forced to happen just from a security viewpoint, big companies won’t be able to get any kind of insurance anymore running AI-infested code.
That last bit needs to hit sooner.
It’s like non-radioactive steel that has to be recovered from sunken warships
it’s more nuanced than that. Claude is made from stolen code, but it generally isn’t going to copy its training data verbatim (unless specifically told to). so copyright wise it’s more grey than strictly wrong. and though claude is made from stolen code, lutris developers are writing something they give off freely to the world, they are not profiting from the stolen code.
does this make it ok? i don’t know. what if they use an open weights model rather than a closed one? would that be more acceptable?
No, open weights changes nothing. Using stolen material is. Especially for a GPL project, a licence normally used to scare off corporate vultures. Why should anyone respect lutris’ licence, when they gave up on the authorship of their own product?
Fork it and call it Ludique, meaning fun in French.
That’s a weird way to run a community facing project, if you want to engage the community that is.
If you treat it like your own personal hobby, you can do whatever you like.
Sincerely… if you can give a single shit about ai in code, you should be able to tell it was used. If you cannot differentiate human from ai authored code, you do not have a seat at the table. jeer from the soap boxes. code is not art. code is code. get over it. does it compile or run and do the thing, cool, fuck cares who or what wrote it. clutching pearls yall cant even define.
People use LLMs to code now, this is not news. Why is claude taking credit in the first place?
Anything generated by an LLM cannot claim copyright, per supreme court rulings. So it is critical to attribute the portions of code that cannot be licensed.
This… is incorrect. Generated code can and has been copyrighted, but not by the model generating it. Humans can get copyrights, digital entities cannot (nor can your pet monkey.) Now, can a human copyright code they did not author? Yes, absolutely. Courts only care that a human had a hand in as little as refining the output or making selections for the agent. Copyright claims look for exercised creative judgement and infringement on existing copyrights.
I had to google “Lutris” to remember what it was. I have it installed… I guess this post made me realize how little I use it and that I should uninstall the slop.
Whether or not I use Claude is not going to change society
This gives me shopping cart theory vibes. I don’t usually base my moral compass based on whether my action will have some kind of measurable impact, but whether I believe it’s the right thing to do. After the intense doubling down in that discussion thread I’m definitely steering clear of lutris. It costs me very little effort to avoid projects that do icky things I don’t want to encourage (even though it may not have a measurable impact~)
Lutris has always been a bit hit-or-miss for me, I avoided it unless it was the only option, as it only worked half the time. I don’t want it to come off like it shouldn’t exist, as stuff making Linux easier to use is great, but I don’t use it at all in my current workflows.
I guess I’ve just been behind the times, but I’ve never had an incentive to switch. I just installed faugus and transferred everything over and it seems very slick. It seems to be missing 1 or 2 things, like environment variables per-game, but all the other important stuff seems to be here. I know what I’m doing with prefixes so having all the knobs to turn is great, but honestly linux gaming does not need most of those knobs nowadays.
How does transferring work?
I only have 2 or 3 things in lutris.
I just did it manually, pointing faugus at the old prefixes and setting the launch options the same
virtue ethics > utilitarianism
Utilitarianism really falls at the first hurdle of any kind of evaluation of a moral system.
It has no real prescriptive power because it demands you be able to correctly foresee the outcome of your actions, something literally addressed by “The road to hell is paved with good intentions”, an adage of at least 400 years ago, and yet people will still gravitate towards it as if society did not explicitly caution us about that mindset forever now.
At this point I can’t help but look down on those who genuinely identify as utilitarian as either too young, too stupid, or actively malevolent and trying to find a way to justify their bad behaviours as errors rather than malice or negligence.
I’d offer you a counterpoint (ignoring the issue with Lutris and AI for a minute):
If you choose not to judge your own actions by the expected consequences of those actions for everyone involved, then how exactly are you supposed to judge them? If you’re following some rule that disagrees with the utilitarian view, then by definition it’s a rule that in your own opinion leads to a worse outcome for everyone.
It’s of course completely fine to not be utilitarian, but trying to claim that all utilitarians are either stupid or evil is just incorrect.
ignoring the issue with Lutris and AI for a minute
Please by all means, I ignored it in the first place, I find this way more interesting.
If you choose not to judge your own actions by the expected consequences of those actions for everyone involved, then how exactly are you supposed to judge them?
Well, this is only half the problem. It’s a bad system because it demands the impossible of you (i.e. accurately predict the future) but it also has a really narrow interest in the dimensions of human morality.
To directly answer the question however: you judge them by a set of principles, whichever you deem right, that you apply consistently across choices.
When it comes to inter-personal choices, the vast majority of all questions can easily be answered by asking yourself “am i betraying some explicit or implicit bond of trust with someone (who has not done so themselves) by doing/saying this?” and if you are, you just stop.
And to be clear, I don’t claim to follow this principle 100% of the time, I am not a saint, but that to me is the guiding principle when there are stakes to my behaviour, and it has not failed me yet.
If you’re following some rule that disagrees with the utilitarian view, then by definition it’s a rule that in your own opinion leads to a worse outcome for everyone.
(Emphasis added)
At its core, the idea of utilitarian morality is to “maximise utility”, that is to do whatever does the most “good” to the highest number of people.
This is, IMO, a terrible metric, and as a deontologist I am perfectly happy reaching a “worse” outcome by it.
It is not particularly hard to see how, by applying this metric, you can justify any kind of scapegoating, abuse, and/or undue leniency on people that would deserve harsh punishment in any deontological or virtue based system, as soon as enough “good” is produced through it.
There is a very dark, but apt, joke about this kind of approach to morality: that 9/10 people involved in it endorse gang rape.
To me, morality is a qualitative assessment, not a quantitative one.
It does not matter how many perpetrator lives will be ruined if they have earned their punishment, and it does not matter how much happier they would be to get away with the crime than the victim would suffer, comparatively.
To do anything else would be to relinquish morality to the whims of the masses, because it implies that there is a threshold past which the abuse of the few becomes negligible due to the benefits it brings to the many.
trying to claim that all utilitarians are either stupid or evil is just incorrect.
To be fair I also stated they can be naïve; I was one too in my youth, until I learned and understood better.
Also, it is one thing to decide that something is not an ethical issue of concern, it is another thing to act with disrespect to everyone with a different opinion.
it is another thing to act with disrespect to everyone with a different opinion.
Unless that opinion is ‘I like using AI’, then they deserved the disrespect.
I can’t fix the problem, therefore I’ll be part of the problem.
At my job we have been told how we have to start using AI more. I can’t really see any point. The only tasks AI can help me for are pointless tasks from HR that shouldn’t exist in the first place. Monthly forms with questions like “how are you feeling emotionally”, used to take me ages to come up with corpo bullshit friendly answers but locally hosted deepseek does it in seconds.
When my work enabled Gemini, I asked it how to disable it. It said it couldn’t help me and asked if I had another question. I didn’t.
That’s the only interaction I’ve willingly had with it.
The HR department will see that it’s not quality human HR-slop and the thought police will be with you shortly
Oh LLMs are great at writing HR slop
In my experience, AI models are fairly good at contextual search. That’s the only thing I use them for.
Yes, if we had documentation then I suspect AI tools could be good for finding information in that.
The Lutris team is small, not corporate, not speed obsessed, etc. I’m inclined to trust them to be among those developers who can use generated code without slopping nonsense all over a code base they know they will probably be stuck maintaining.
