this post was submitted on 03 Feb 2024
663 points (90.4% liked)

linuxmemes

20880 readers
6 users here now

I use Arch btw


Sister communities:

Community rules

  1. Follow the site-wide rules and code of conduct
  2. Be civil
  3. Post Linux-related content
  4. No recent reposts

Please report posts and comments that break these rules!

founded 1 year ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 1 points 9 months ago (1 children)

Fantastic breakdown and rebuttal. I concede on all points save one. But before I nit pick a bit a couple things.

One is thank you so much for tsking the time to rebuke me with such detail and finesse. I enjoy prodding people on these things because I find back and forths like this more engaging and informative when a stance is taken rather than just Q & A. I know it must bave taken you a bit of time to write up thede answers, so thank you.

Secondly, I just wanted to commend you on totally eviscerating me on the Only Office bit. i had forgotten on that point and hadn't taken the simple step of searching beforehand. It was a poorly made point and I sincerely apologize for posing that weak argument. Yeah yeah, you're probably thinking "the entire argument was weak". And I'll not try to convince you otherwise (I'd be unlikely to succeed anyways, right?).

Now, my only point of contention:

No we dont need an alternative always. We dont need 6 audio recorders, we need 1 good one that does everything right, is fast, secure and usable.

Now, my only contention here is that competition, true competition, is good. I'd say you always need at least 2 major nearly equal players in any of these fields. I don't want there to be 1 Linux distro, I don't want there to be 1 Office Suite, I don't want there to be 1 package manager, and hell, I don't want there to be 1 display protocol (and for that I am very happy wayland exists for that reason alone).

Competition over best implementations is good, and more selfishly for me in particular, more choice is good. You can argue that those choices can stifle innovation as it divides the talent base over possibly trivial minutiae of implementation (or just create a poor implementation outright), but ultimately what drove me to Linux was not my admiration for it being secure or light weight, but rather it is the availability of the many choices available.

I'd rather not see that wrangling up of the diversity that exists within the Linux ecosystem go away in the interest of conformity to a singular best practice. With all the consequences that entails.

[–] [email protected] 1 points 9 months ago

You are welcome, it took some time and I actually accidentally swiped back and deleted like 6 paragraphs but hey.

Now, my only contention here is that competition, true competition, is good. I'd say you always need at least 2 major nearly equal players in any of these fields

I dont know... KDE is using Qt, GNOME is using GTK. KDE breaks all the time, GTK attracts many developers of small software with nearly no customizability, but that works.

KDE apps are still looking a bit dated but refuse to follow the "padding everywhere" BS that GNOME, Windows 11, MacOS etc use.

I dont know what the word "competition" means in FOSS. These are not companies serving customers, fighting for marketshare. These are just products by and for the community. It sucks that KDE does all the cool stuff, but is inherently memory unsafe as f*ck, GNOME not even having the most basic features but being very stable, and nobody caring about Cosmic really.

Wlroots is nice, and it would be really cool if all projects could just use that. Wlroots is not complete like KWin.

Actually, XOrg, Linux, GNU, there are so many projects that just dont have an alternative and that helped to create products that all work but have a different look and feel. Under the hood they where all just fancy XOrg.

I think there are problems with monopole projects that are bloated and eat up more and more subprojects. This makes 0% sense and should not be done.

The Linux kernel is a mess. It is full of random vendor blobs for XYZ hardware, poorly written code (according to Jeremy Soller) and everything on every machine.

Look at windows. It kinda "feels weird" to have those branded "AMD Radeon Driver®" display in the task manager. But the fact that they show up, nobody gets that. They are seperate processes in the equivalent to Linux userspace. You can restrict them, give them permissions etc.

It "just works" but its horrible. Any random code by any weird manifacturer just gets thrown into the Kernel, because Distros can't unite on how userspace is supposed to look like. So instead of fixing that problem and putting all drivers into userspace so users can just use what they actually need and just remove the rest, we have this huge and not even FOSS blob that runs everywhere.

I think I want to switch from Fedora Kinoite to something like NixOS, as I think building the kernel for your actually used hardware, removing everything else, is essential for security.

The next project is systemd, which works well, is somewhat nice to manage (I still find it very confusing to create services but I guess this is nice?) But it is pretty horrible.

It is a huge binary, a single one, always running. You would need to fork it and remove and replace stuff to not break it. It is a de facto standard and makes no sense, why would you

  • bundle everything in a binary
  • make it impossible to replace parts
  • use a memory unsafe language with no sign to switch

Their Github issues are insane, I cant imagine anyone even wants to look at 1k open issues. This would simply not be the case if it was split up. Could still be used as a bundle, but if (like with rust rewrites of GNU core utils, like uutils) people would rewrite parts, they could test them seperately and slowly move the project to Rust for example.

So yeah, monopole projects suck if they are not modular. Desktops should work on the same things though, to make them work well. GNOME is supposedly very specific about mutter so nobody wants to use it, but KDE could possibly switch to wlroots if it has feature parity with Kwin and that would really reduce useless duplication of work.

You dont need competition, just talk to people in the same project, everyone has different goals.