Asahi Linux To Users: Please Stop Using X.Org

[mrg666]

The HPC systems I use at work only have X;org. My personal Linux workstation at home is configured to work with Wayland. So, I do Wayland for locally running desktop apps, X11 for remote apps running on a server over a ssh tunnel. I think both will be available as long as needed.

[microcode]

I'll stop using it when Wayland is at all usable lol. Funny for a project with a Japanese name to suggest that you use a display server environment that still utterly fails at consistent input method support, especially if you are running anything except bog-standard GNOME.

[qarium]

Well, it depends on what you want to do with it. If we are talking about notebooks based on relatively low performances SoCs (mainly repurposed older smartphone or setup-box SoCs) then there are quite a few options. This one, also, Chromebooks and several Windows on ARM machines. They are OK for small tasks like web browsing, terminal emulation, basically, for the thin-client use cases. I had in mind specifically high performance SoCs, which would be capable to provide comfortable experience compiling large projects, running several docker instances etc.
Also, personally anything below ARMv9.0-A I consider to be legacy at this point (in the context of a high performance ARM notebook I mean). I was already not too happy that M1/M2 is around 8.5/8.6, but If Apple is not going to bump M3 at least to 9.0A (implies SVE2 support), I will be a bit disappointed

right. of course i know the rockchip 3388s is slow in comparison.

https://gadgetversus.com/processor/apple-m1-vs-rockchip-rk3588/

as this comparison shows the difference is huge.

of course ARMv9.0-A+SVE2 will become really a awesome tech.

not only that we can expect 3/4nm ARM SOCs to in the near future. some ARM cores will be at 3,4ghz compared to the 2,4ghz of the rockchip this will give a massive boost.

[Bounding1662]

Have these people on Asahi Linux been living under a rock or something? Because if they didn't then they would know that Wayland has still got serious problems even nowadays where development for it has accelerated? I really don't understand this sentiment. Yes, we too would like to just switch to Wayland, but guess what, we simply CAN'T, the project simply isn't ready yet

[Quackdoc]

Having to use car analogies in a software discussion is never a good sign .

why?

I used to use GNOME(bad idea) on my laptop between 2020 and 2021 due to it's supposedly better touchscreen support and it had a bug that made it froze completely if you made the mistake of dragging an icon from the dock to the desktop area, a quite annoying but for a 2 in 1 user, I encountered it frequently and the only solution was restarting GNOME. Did this mean that GNOME was broken?

if it's an issue that effects a lot of users and breaks usability then yes

No, it was buggy. A broken program doesn't achieve its main function at all just like how a broken watch doesn't show the correct time randomly for some people(maybe it does two times a day but that's beside the point). Argue semantics all you want if you want to waste your time but you won't convince me.
...
And stuff like this is coincidentally a very good example of why I look at claims that KDE is broken or otherwise unusable as hyperbole. I've been using my laptop for notetaking with xournal(fullscreen and otherwise) for quite a long time now and never had this issue(or a problem with rotation in general).

Ok and? who cares what you have, once again, I reiterate, works for me so it must for thee, anyways here is the issue ticket. it's not hyperbole, kde has an entire issue tracker chalk full of issues like these. not sure why it's so hard for people to understand that just because you don't have the issue, doesn't mean other people dont. https://bugs.kde.org/show_bug.cgi?id=467138

Guess which DE I am using. Convincing someone who has never had a problem with X that X is broken and unusable is not a very easy task for reasons that should be obvious. Anyway if we are labeling things as broken due to issues of that severity then GTK was broken for me for over 18 years due to not having thumbnails in the filepicker. For some reason I'm sure a lot of people wouldn't call that being "broken" but who cares? If I see a bug then it is broken.

X is broken, but wayland is broken too. the issue is that for tons of people, wayland is still competely and utterly unusable, so the quick force to wayland, when it's not ready, will wind up pushing people to other operating systems like OSX or windows.

I do not think your premise is valid. Corporate use typically does not involve desktop environments and when it does, X11 and Wayland are both equally good. Please do not reply with fallacious reasoning about RHEL’s choices. RedHat just wants less work to do and that means picking only one protocol to support.

Alan Cooper and others are already maintaining the Xorg X11 server. My point is that no amount of maintenance will be counted as maintenance by its critics here. They will just spout nonsense about it being unmaintained.

it is valid, the issue is that it is very much an issue. No on wants the burden of maintaing X. the people who are maintaining it don't want to. the only reason it has taken this long is because they still deem X necessary for one reason or another. corporate is what is sustaining X. I have no doubts that within the next few years other corporations will start dropping X support. the only question is if wayland will be in a state to fill the needs of people who would otherwise be left behind

[ssokolow]

Another annoyance - applications cannot remember and restore their positions. Maybe it's "by design" but if so that's a really silly decision, and I just can't relate as a user.

That's actually one of the reasons why, as a developer, I migrated from PyGTK to PyQt4 back in the day.

QMainWindow comes default with toolbar and panel customization and, paired with QSettings, it's six lines of code to remember and restore both window geometry and toolbar/panel customizations. (That sort of "GTK requires you to reinvent it from scratch; Qt comes with it built-in" pattern repeats in all sorts of APIs.)

[fitzie]

Wayland is broken for ecosystem reasons, not technical reasons of which there are plenty. The DE are burdened by their legacy code, and NIH approach. The incumbents are tied to certain technologies, like UI toolkit or preferred programming language. Proponents of wayland should go after the DE maintainers for not using wlroots, the toolkit writers for not allowing compositors to restart without crashing all applications, and the remote desktop frameworks for not figuring out how to remote display a wayland app. Saying wayland is the future so you'd better just accept is, just shows how much apathy there is towards people who really just love their linux desktops. I'm totally on board with wayland, but I cannot say it hasn't been a mess.

[WiR3D]

I'll stop using it when Wayland is at all usable lol. Funny for a project with a Japanese name to suggest that you use a display server environment that still utterly fails at consistent input method support, especially if you are running anything except bog-standard GNOME.

Lol. I'm running an AMD kaveri htpc on wayland with a heavily customised GNOME. Haven't had a single issue since ubuntu 20.04. Only thing that gives me shit is kodi and vaapi regressions.

[WiR3D]

Being right... doesn't make how they said that good.

Also its hysterical that they call Xorg broken garbage... because Wayland is literally built on top of the same drivers. Xorg is no more broken today than it was 10 years ago. Saying thing like that is shooting around thier own feet.

Lol you obviously read nothing in this thread or Hector's mastadon Threads. Basically all x11 devs are now on wayland and only fixes being done to x11 are CVE'S, and if you are reaaally lucky a bugfix

[WiR3D]

Same for Wayland, I never expect those issues to ever be fixed neither.

Very strong words, have you by any chance seen the number and quality of the discussions on the wayland protocols issue tracker? Or the kernel patches and designs that the wayland team has had hands in designing? Me thinks not.

The amount of things they are improving ecosystem wide is staggering, truly. But yes I'll concede there is yet more work to be done.

For you though, the problem is probably in the compositor, or setup (barring Nvidia which is still a debate raging in the kernel dri-devel mailing lists, not in wayland fyi)