>>42492720
That's okay, anon. I even agree with you, Linux a shit - but unfortunately so are all other systems I've used so far, in their own ways. I just think people deserve a chance to try every OS if they want, without prejudice that can be gained by reading someone shitposting for lulz. One man's garbage is another man's waif-, uh, treasure, after all.
>>42492730
>I don’t think that makes me a gatekeeper
Eh, it's a bit more nuanced than I can fit in 3000 words, so pardon me anon for making it seem like I implied that.
If memory serves, Steve Jobs once said that "if an average user is required to use commandline to do their task, you've fucked up as OS UX designer" or something to that effect, and I mostly agree with that sentiment. You probably have seen how often people online tell other people to "just run those commands to fix problem", and how people run them without understanding what they'll do. `curl | sudo sh -` being so popular continues to baffle me.
I can, however, also believe someone being so tired by searching a solution to the problem (how 2 steam on popos), so that they will try even the unknown commandline and glance over/not fully understand warnings like these. If the package solver in apt needs to support actions where installing package A removes packages B, C and a lot more, such operations should have at least a second question to confirm that you do understand what you're doing. And if the user isn't a nerd, they should not be encouraged to use lower level tools without as much guardrails as you'd put in GUI. But you know, online websites (and now LLMs, yay!) promising solutions that maybe worked on some flavour of Linux at some point of time. Maybe it's just me, but I can't blame the stumbling user significantly more than everyone else who led to this painful experience.
Not that I'm happy about how software is packaged on Linux systems in the first place. For an average person, I believe the model found in Android might be overall best. Flatpak seems closest to this so far, but as I've said it needs more time (e.g. replacing static permissions with dynamic), and as always in engineering, there will be tradeoffs. Maybe some other tool will emerge, who knows. On that note, ChromiumOS design documents are quite interesting in regards to the less common userspace they've got.