I wasn’t sure if it was a joke and I read another comment here that seemed to indicate that it was not. Thanks for clearing that up, unfortunately sarcasm is easily lost across the internet.

No clue what your issue is, but if your standard for OSS is not being associated with furries, I’ve got bad news for you.

Depends on your use case. Manjaro is in the unique position of having such a poor management track record, that almost any other Arch derivative (or, you know… Arch) is preferable. They’ve let their SSL certificate expire for the 7th time just last week I think? I honestly haven’t kept count, but it happened at least a few times. If it works for you, that’s fine. I imagine most reasonable people don’t really care. I personally dislike Manjaro for their poor track record managing their repositories or SSL certificates, their historically stupid approach to the AUR and the generally negative effect it had on Arch linux (tech support). Honestly I think Manjaro benefits most from its popularity in the tech influencer spheres and the fact that most of its userbase doesn’t seem to question if a rolling release model actually makes sense for their use case.

I mean, yeah that would be my solution. I get that the AUR is attractive, precisely because it has a low barrier for anyone to submit their PKGBUILD. The level of oversight and verification is just a bit too low to recommend it to an average user, without a lot of caution. You’ve mentioned some alternatives that fall on different points along the spectrum of delivering software. Something like flatpak is a much more reliable tool in the hands of someone who just wants a GUI app and not think about how it gets to their desktop. For everything else that isn’t part of your distros repositories, there’s really not a good noob-friendly solution that doesn’t carry a big potential risk. Most distros have third-party repositories that use the same underlying tools to deliver software, but are less strict about QA and stuff. This is kind of a bad fit for rolling release distros in my opinion and is probably one of the reasons the AUR is so hands-off and DIY oriented.

There’s probably a better way to handle this, but I don’t think it’s an easy thing to solve (especially for the rolling release model) and the AUR isn’t really appropriate for mass-consumption by average users. Also, there will always be a certain point beyond which you’re on your own, it’s just not feasible to have reliable, safe, distro-agnostic packaging for every piece of software out there.

All official resources, Arch maintainers and high quality guides have been putting a ton of effort into teaching people how to use the AUR safely. That hasn’t stopped some people, even back before Arch got really popular, but you can’t reach everyone. Alternative package managers and pacman wrappers made the AUR a lot more accessible, which isn’t necessarily a bad thing, but there are good reasons for all the caution. Combine that with Arch increasing in popularity and getting picked up by all the shitty influencers and you get a lot of people ,who don’t know what they’re doing, installing everything from the AUR with their CLI/GUI of choice. Then you’ve got Arch derivatives making AUR packages easily accessible from the start, bad advice on places like reddit etc.

Long story short: it seems that over the years whenever I check in, users that barely know how it works are happily installing random shit from random people on the AUR because they saw it in a YT video or something.

The AUR is a great resource but it’s also being sold as a package repository users don’t need to actively think about or understand. I honestly think malware is going to be much more common on the AUR if we aren’t careful.

Microsoft has been weirdly protective of their installation image downloads for some time now. I don’t really get why.

I’d be fine with it, but I had to pin the flatpak to an old release for now, because the redesign fucked up the migration, deleted all my presets and also caused issues with voice chat apps not getting any audio input. The new UI is also unquestionably a downgrade and less accessible when it comes to setting slider values, but I hope that can be fixed with time. I certainly don’t blame FOSS devs for their work.

Btw. seeing some of the comments in here: is it the fate of all Linux shitposting forums to be filled with hardliners who really care what software you install on you Linux system? Let me use my GTK apps in peace. I don’t need opinions on UI cleanliness and density from people who don’t even use easyeffects, because my god, is it a mess currently.

Last time I used Nemo it was an unstable mess, but that was a while ago. I’m sure it’s been improved.

The Lunduke effect.

It only checks mirror latency every transaction if you enable “fastestmirror”. Repositories are only synced if the local cache is out of date or the metadata timestamp has changed. There might be a way to prevent dnf from refreshing repository metadata at all, but I really don’t think that’s a good idea.

In my experience, dnf has pretty good mirror selection by default. Setting “fastestmirror=true” replaces the more complex mirrormanager2 heuristic, which tries to select an appropriate mirror by available bandwidth, with a simple latency check that runs before transactions. In most cases this has no effect or worsens dnfs performance. They changed the description in dnf5 to better reflect the behaviour.

Having said that, it’s worth giving a try in a case like this. I just want to make sure that people realize that there is a reason this was never enabled by default, since this is a popular configuration tweak suggested all over the internet, whose actual function very few seem to know.

Good marketing or influencers perhaps? I’m not sure but I’ve been wondering myself.

If you use omarchy, you have bad taste.

The default paste action is pretty much the only thing preventing anyone from picking a different function for the button. That’s the the biggest reason for reversing the default behaviour.

There are programs that use the middle mouse button but also support pasting from clipboard. I’ve been annoyed at work plenty of times when I’m trying to translate across a canvas but accidentally paste a random node of text. Bonus points if it contains some kind of password that was still in your clipboard. I don’t think it’s a good default.

The choice of programming language for new projects almost always comes down to the existing ecosystem and its popularity in the space you’re working in. Some of the most popular languages also have the advantage of being deeply embedded in the academic pipelines educating newcomers to the field, which obviously includes C++. Being one of the worst choices for a first programming language to learn hasn’t stopped C++ yet and likely won’t for a long time to come.

Asking for help online just gets you a “lol, RTFM, noob!”

Depends heavily on what place you ask for help in. There are plenty of spaces explicitly meant for community tech support. In OPs case, I’ll say the title doesn’t help and asking an LLM for advice on a topic you’re unfamiliar with (and not second-guessing the commands you paste into the terminal) is such a bad idea that it really can’t be understated. I regularly catch some of my colleagues making AI-assisted mistakes and they’re professionals who genuinely know better. This shit shouldn’t ever be recommended as a learning tool for beginners without some kind of supervision or guard rails to ensure you’re not being gaslit.

I like the '89 and '93 Patlabor movies.

KHTML and WebKit is a historic mess but it’s debatable at best if Apple actually violated license terms. In any case, it shows just how ineffective LGPL is at enforcing the intended contributions from corporate licensees. I’m not getting into this historic mess of a topic with someone who has yet to give a reason why Rust needs to be singled out for being MIT licensed when it was already the de facto default choice for most open source projects before it ever became popular. It’s quite clear to me from the endless brain-dead comments in Lundukes YT channel or in the Phoronix forums, that a vocal minority of the Linux community has a massive hate-boner for Rust and is desperately trying to come up with a valid reason for it. None of these people are actual experts from what I can tell, but boy do they have strong opinions about the programming languages used by the people who do all the work.