Latest 30 Comments

Quoting: pbMy son is on manjaro and his system breaks on every update. But apparently arch is "too complicated". 😆

You're exaggerating, don't you? ^^
Like many others, It works pretty well for me and I'm using it for 10+ years, so pretty stable in my eyes. Usually when it breaks it's my fault, not really Manjaro's. I'm on XFCE, it seems to be the most stable of all, the KDE folks did have some issues more than once though.
I hope positivity comes out of this, I'm happy with Manjaro and don't plan to change right now.

View this comment - View article - View full comments - Report this comment

Quoting: TriciaPearson

Quoting: dragonherder He did not initially hide it though. He hid it after basically being bullied gor using AI to finish work he hasn't had the motivation to do on his own codebase... he then got bullied for removing the attribution because now he was trying to hide it. As another pointed out and I responded to it is basically coming down to people bullying developers using AI no matter what

That's also partially untrue. At the very first he didn't hide it, BUT he did hid it BEFORE being bothered. See his own statement :

" No, I removed the Claude attribution way before anyone complained. Looking at the git history, it seems to have happened on Feb 18th. The only thing I did as a result of this thread was putting them back. As a bonus, it fucks over every news source saying I've hidden Claude usage, making their articles wrong and outdated. "

Likely something aimed at us for reporting on it. But it goes to show how confusing he's made the situation by hiding it, and his later comments about hiding it. The whole situation is a stupid mess of his own creation.

Personally, I've removed Lutris and won't be using it myself or suggesting it in future.

View this comment - View article - View full comments - Report this comment

I love the idea of Manjaro - Arch, but with milestones (stable releases). It sounds like the power of Arch, but without the constant fear that every update will bring issues. And pamac - lovely. They've actually addressed one of the biggest issues (for me) of Arch.

But now that I'm on CachyOS with its built-in Snapper/BTRFS integration, I have that comfort anyway. If I do an update and my system has issues, I just reboot and choose the snapshot before the upgrade via Grub. Easy.

I still miss Pamac though. I haven't looked to see if it's portable to a generic Arch install.

I hope Manjaro get their organisational ducks in a row. It would be great to see what they could achieve without constantly having to fight internally with each other.

View this comment - View article - View full comments - Report this comment

My experience with Manjaro had been pleasant back then - there were some issues with libcrypto migration and some updates had had some issues, but never something more than a quick snapshot rollback + reupdate wouldn't fix (well, aside for an issue after 8 month of not upgrading as my PC was left at home but that's just a known Arch issue).

What made me stop was the way they removed, restored, and then removed some codecs without adequate dialogue and then not mentioning it in the updates announcement. Between that, and the various management mishaps, I just chose to move to Universal Blue, then Bazzite, and now CachyOS.

I think their community is right in wanting to break away, though I would say making a new distro without the associated branding of Manjaro that is just toxic now.

View this comment - View article - View full comments - Report this comment

I've had the same install of Manjaro on my main PC for over 10 years (throughout a system rebuild with new motherboard). I've had a couple of breakages, but nothing insurmountable, overall its been pretty stable... (TBH, I did try to move off Manjaro a few years ago, but had some weird issue with the calamares installer which all the distros I wanted to use employed, so I just ended up staying on Manjaro)

View this comment - View article - View full comments - Report this comment

Would suck to lose Manjaro. It's basically Arch, but with the Stable branch you can update every time when there is an update. You don't have the same sort of commodity with Arch where with updates stuff can easily break so you have to postpone updates and look info from packages a lot.

Same goes with CachyOS. There can be breakage happening on those much more easy.
Something that is Arch and postpones updates and have testing for those is much wanted IMO.

Ive been basically years with Manjaro and it has been rock solid for me. Would be sad to see it go.

View this comment - View article - View full comments - Report this comment

That's interesting. I've been using it for over a year and haven't had any technical problems. That doesn't negate the organizational problems they clearly have though. It sounds like this change will be for the best, assuming it survives at all. Thankfully, there are plenty of Linux choices out there, and it's pretty easy to move on to another.

View this comment - View article - View full comments - Report this comment

My son is on manjaro and his system breaks on every update. But apparently arch is "too complicated". 😆

View this comment - View article - View full comments - Report this comment

Quoting: kmturleyIt's sad to say this, but you will need a 100% human badge for the site.

Also sad to say that I genuinely need this to be a thing these days.

View this comment - View article - View full comments - Report this comment

Sounds like 100% marketing bullshit to me ... There's been native Linux support from Unity for at least 12 years now (we've been shipping our Unity based game natively to Linux since back then without any major (not homemade, like shader fuckups) issues. Including Steamworks and whatever other "plugins" we needed. And - as that was mentioned - 100% modding support parity to the Windows build, including code modding, both by adding new code as well as patching existing through Harmony(X).

While I could see them improving the *Linux* target in general (like was mentioned the Vulkan renderer still isn't perfect, at least in 2022.3, as it's way more dependent on the HW/SW combination of the system than both the GLCore and DirectX renderers). But why should we ever have to make different builds for e.g. "Steam Machine" vs "Linux"?

View this comment - View article - View full comments - Report this comment

I used it a couple years ago to install Ubisoft Launcher, but my gosh it's clunky AF, and making it "sloppier" isn't the right path.

View this comment - View article - View full comments - Report this comment

Quoting: Arten

Quoting: CaldathrasThe misnomer "AI" is just a marketing label used to inaccurately promote LLM. In my opinion, when the backlash began against LLMs and their objectionable social, legal, economical and environmental impacts, it was convenient for the spin doctors to confuse the issue by throwing translation, grammar checking, spell checking and other older technologies that pre-date LLMs under the same label. How could we object to these uses of "AI"? We don't. We object to LLMs and how they are being promoted and presently utilized. Personally, I also object to the label "AI" because LLMs are any but intelligent.

My two cents.

Thanks, Liam, for your determined stance.

Spell checking alone isn’t enough. As a Czech speaker, I find that having an LLM correct Czech works much better. Czech is a stupidly complicated language with many illogical exceptions, so we at least need the ability to guess the context for a correct correction. Spell‑checkers were terrible before LLMs.
The same goes for English when you can’t formulate a sentence correctly. My English is terrible, so using an LLM to correct it is very useful. Can a spell checker do that for me?

There's nothing wrong by using AI as a TOOL! But you woulnd't like to be replaced entirely by it and let it post on this site in your behalf... I guess.

View this comment - View article - View full comments - Report this comment

Humans (and Penguins) are the only species allowed here. 😁

View this comment - View article - View full comments - Report this comment

Quoting: scaine"I'm doing a thing people hate, so instead of not doing that, I'll continue to do it, but hide it better"

That's a bold position to take in any project, let alone a FOSS project. And just because he can use it in non-slop manner (maybe? hopefully?) it doesn't absolve him of the ethical concerns many of us have about genAI.

There are no ethical issues worth our concern. We'll carry on, not caring about your feelings.

View this comment - View article - View full comments - Report this comment

Quoting: CaldathrasThe misnomer "AI" is just a marketing label used to inaccurately promote LLM. In my opinion, when the backlash began against LLMs and their objectionable social, legal, economical and environmental impacts, it was convenient for the spin doctors to confuse the issue by throwing translation, grammar checking, spell checking and other older technologies that pre-date LLMs under the same label. How could we object to these uses of "AI"? We don't. We object to LLMs and how they are being promoted and presently utilized. Personally, I also object to the label "AI" because LLMs are any but intelligent.

My two cents.

Thanks, Liam, for your determined stance.

Spell checking alone isn’t enough. As a Czech speaker, I find that having an LLM correct Czech works much better. Czech is a stupidly complicated language with many illogical exceptions, so we at least need the ability to guess the context for a correct correction. Spell‑checkers were terrible before LLMs.
The same goes for English when you can’t formulate a sentence correctly. My English is terrible, so using an LLM to correct it is very useful. Can a spell checker do that for me?

View this comment - View article - View full comments - Report this comment

I don't give a fuck. Just as long as it works.

View this comment - View article - View full comments - Report this comment

It's still Unity... the next bait & switch is coming.

View this comment - View article - View full comments - Report this comment

Quoting: Philadelphus

Quoting: Liam DaweOpenTTD has had free graphics and music available for a very long time. Using content from TTD is entirely optional.

That's what I thought. This definitely smells of Atari going "make people buy our game to play yours, or we'll DMCA you off of Steam and GOG".

Oh, I don't know about that. It's certainly plausible but not the only explanation. The OpenTTD team may have chosen to do so preemptively to avoid any potential complaint from Atari. Another possibility is that Steam & GOG may have asked them to do so, as well. The norm with these open-source executible projects (and total overhaul mod bundles) has been to offer them only to those that own the original game on the same store. This change puts OpenTTD in line with that approach.

View this comment - View article - View full comments - Report this comment

The misnomer "AI" is just a marketing label used to inaccurately promote LLM. In my opinion, when the backlash began against LLMs and their objectionable social, legal, economical and environmental impacts, it was convenient for the spin doctors to confuse the issue by throwing translation, grammar checking, spell checking and other older technologies that pre-date LLMs under the same label. How could we object to these uses of "AI"? We don't. We object to LLMs and how they are being promoted and presently utilized. Personally, I also object to the label "AI" because LLMs are any but intelligent.

My two cents.

Thanks, Liam, for your determined stance.

View this comment - View article - View full comments - Report this comment

I just stopped reading most sites. If the article, reads like it's written by an eighth grader, I just stop reading and never go back to that site again. Same with YouTube. On YouTube, I don't allow a history to be kept. That means there are zero suggestions. The only way I have videos to watch, is through people I follow or searches. If you are a real person, do everything you can to demonstrate you are not AI.

Maybe we should have an AI Tracker list. Up till now, I've just been removing them from my favorites and forgetting about them. Maybe I should have been adding them to my pihole block list.

View this comment - View article - View full comments - Report this comment

Quoting: Salvatos

Quoting: CaldathrasYou are assuming that most people even care or know what operating system is on the device.

You’re putting words in my mouth. What I’m assuming is that if their devices annoy them, they will complain to vendors, and the vendors will lobby against the law when the customer service it requires or the bad press it generates become too much of a burden/liability.

Not at all, you did write

Quoting: SalvatosLinux runs virtually everything that's not a PC.

I simply pointed out the most people are not going to care about the underlying operating system on those devices. The discussion, after all, was that Linux and Linux alone would inspire a voter uprising.

I'm not saying that a voter uprising won't happen, just that if it does, it won't be Linux that inspires it. They won't care about the operating system, just the inconvenience the law is creating in their lives.

I can't see complaints to vendors initiating such political change. If the vendors had that much influence, the bill never would have been tabled in the first place. I don't think the vendors will care -- such laws are rarely retroactive.

Myself, based on the acceptance of the Microsoft Account requirement for Win11 and everything else the Tech Giants impose on them, I'm guessing that the California voters will just roll over and accept it. I could be wrong.

Do voters in California have the ability to initiate a mandatory referendum, that, if successful, the politicians are required to implement?

View this comment - View article - View full comments - Report this comment

Quoting: scaine"I'm doing a thing people hate, so instead of not doing that, I'll continue to do it, but hide it better"

That's a bold position to take in any project, let alone a FOSS project. And just because he can use it in non-slop manner (maybe? hopefully?) it doesn't absolve him of the ethical concerns many of us have about genAI.

If you don't trust AI generated code is just because, maybe, you put too much trust into humans and that is dangerous, as the very opposite.
As for the ethical matter, do you think it poses an ethical issue to write an official document using a text editor, instead of doing it by your hand?
This is the same case, except the "co-authored by AI" sign thing which is very stupid, indeed, and without any legal basis.

View this comment - View article - View full comments - Report this comment

Quoting: CaldathrasYou are assuming that most people even care or know what operating system is on the device.

You’re putting words in my mouth. What I’m assuming is that if their devices annoy them, they will complain to vendors, and the vendors will lobby against the law when the customer service it requires or the bad press it generates become too much of a burden/liability.

View this comment - View article - View full comments - Report this comment

I played the game as a demo about a year ago, and I like it. I've been biding my time, waiting for it to be reduced in price, maybe 50%+ ideally 75%+ discount. Although an unrealistic viewpoint, I wish developer teams would *move on* to new projects rather than keep adding extra 'value' onto older titles. DLC additions like this keep the asking price of the title high which discourages me to ever buy it.

Yes, I am a proper bargain seeker. It is almost unheard of for me to pay anywhere near full price for a title.

In a similar vein, rather than the Facepunch developers keep on changing their RUST game, adding more and more content, completely altering the way it plays - They could have built a brand new game from scratch, survival, base building, in a different setting (Space, Fantasy, Cyberpunk).

View this comment - View article - View full comments - Report this comment

Quoting: mr-victory

Quoting: ArehandoroThe fear is tangible.

And fear is not good for working.

The fear of what? Merging so much code with unknown functionality and/or side effects?

That too, of course, but mainly fear of being let go. If AI writes the code, and AI reviews it... what are we supposed to do?

View this comment - View article - View full comments - Report this comment

Quoting: ArehandoroThe fear is tangible.

And fear is not good for working.

The fear of what? Merging so much code with unknown functionality and/or side effects?

View this comment - View article - View full comments - Report this comment

Quoting: tuubiThe only way I've used LLMs professionally is to help with the translation of documents I've written, and that's only due to time pressure. It helped less than you might think, seeing as we're talking about compliance and policy documents, which means the language needs to be precise and exact. Took a lot of manual comparison and corrections, but it was still a bit quicker than doing it all line by line.

I think that is the key point. Most people have tight deadlines and end up using it, myself included, even if they hate it and are against it.

In my team, a mix of software engineers, architects, data engineers, etc, we're now expected to use Claude Code, and from creating code and architecture design now the workload has shifted to reviewing PRs... and only last week the company started playing with AI PR reviews too because they became the bottleneck, cause people can't properly review so much slop all the time. The fear is tangible.

And fear is not good for working.

View this comment - View article - View full comments - Report this comment

Quoting: Arten

• AI is just a tool, and when used responsibly it is reasonable to employ it. I don’t think generating an entire article with AI is a good idea, but using it for spell‑checking? If it can catch incorrect wording, that’s a worthwhile use.

This is the correct take, as long as "responsibly" and "reasonable" are well defined, and cover issues beyond immediate utility, including legal, societal and environmental ramifications. Not just whether you get useful results out of it.

The only way I've used LLMs professionally is to help with the translation of documents I've written, and that's only due to time pressure. It helped less than you might think, seeing as we're talking about compliance and policy documents, which means the language needs to be precise and exact. Took a lot of manual comparison and corrections, but it was still a bit quicker than doing it all line by line.

Mostly LLMs have been a huge pain in my backside and a source of tons of extra busywork, being the guy responsible for information security in our organisation. I keep having to preach to everyone from the C-suite down to individual developers that laws do in fact limit what we can do when it comes to LLMs, and risks need to be assessed and results measured just like in everything else we do. We've got EU wide legislation like the EU AI ACT, NIS2 and obviously the GDPR forcing us to be responsible with our customers' data and transparent about our AI use. (And that's a good thing for every single one of us, even if that might mean a few less euros going to the pockets of shareholders.)

And non-professionally, I don't see any use for LLMs. I enjoy being creative, be it photography, coding or doodling something silly with my wife's Cintiq, and all of these would be less fun if the machine did the work for me. Results are less important than the process. Might as well just let an LLM play my games and read my novels for me.

View this comment - View article - View full comments - Report this comment

Quoting: kmturleyIt's sad to say this, but you will need a 100% human badge for the site.

The idea is so good that I was sure somebody has made them already.

Here you go: https://www.aimeecozza.com/human-made/

View this comment - View article - View full comments - Report this comment

Thank you for making this site my #1 bookmark on everything! The work put in daily is nothing short of incredible.

View this comment - View article - View full comments - Report this comment