Via the sidebar at <https://www.reddit.com/r/kde/>:
Plasma/Wayland Known Significant Issues - KDE Community Wiki
community.kde.org
oh... ooh, oooh... How come Wayland issues can be traced to NVidia drivers, but not a single mention of Wayland issues getting traced to AMD (Via the sidebar at <https://www.reddit.com/r/kde/>:
Plasma/Wayland Known Significant Issues - KDE Community Wiki
community.kde.org
amdgpu
) ? Especially considering I got Wayland working on an RX 6900 XT ? … Wayland issues can be traced to NVidia …
… please, how might we achieve the equivalent of <https://wiki.archlinux.org/title/NVIDIA/Tips_and_tricks#Preserve_video_memory_after_suspend> on FreeBSD?
To be fair, neither have many Wayland compositors. Most of the work is in the mesa / libdrm / modesetting layers.X.org hasn't gotten any major updates in ages anyway, so it'll always be there for those who prefer it.
Wayland lacks screen saver ability, reparenting ability which some want, and a good way for server client forwarding. The screen saver is an important one for everyone.
But it supports screen locker? Why not just rewrite a compatible screen locker to behave like a screen saver? (I'm just throwing an idea out there, I have no skill to actually do that). It's entirely possible that this idea is not technically possible due to the very design of Wayland...Wayland lacks screen saver ability
For something that is pretty much feature complete
… what I read (and linked to earlier) …
Maybe we need to jump all over Wayland for those things.
… There’s a bit of chuckling and jeering over this in developer circles, …
old??? funny, graphics/blender (since 1994) is still 'new kid on the block', even though it's arguably just about as feature-complete as AutoCAD (since 1982). Nobody calls either of those two 'old' or 'crappy'.The kids nowadays call that "old".
So many times people say something must be bad just because it's been around a long time. They never consider that maybe a thing is just done.
Additional discussion:
Truth be told, the best outcome would be if upstream Wayland developers would finally acknowledge that there are application and desktop environment developers and end users out there who don't buy into this whole Portals-Flatpak-Systemd-Pipewire thing, and deliver feature parity with X11 on "pure" Wayland. If that happens, mission accomplished!
Yeah, Wayland is about as incomplete (in comparison to Xorg) as FreeBSD is (in comparison to Windows).Wayland is incomplete, and X.org is riddled with vulnerabilities; which endeavor are we willing to alleviate if we were to forge our own path?
And then there's armchair analysts like us. We don't really know the technical difference between Wayland and Xorg beyond a bit of superficial research, just bringing up random examples from the Internet and holding them up as definitive and authoritative evidence that one is better than the other. This is at least partly why devs who do the actual work on Wayland have such disdain towards uninformed noobs who pretend to know everything needed, and hide behind the 'customer is always right' idea. Well, if you know your manners, and are aware of your position, you can get surprisingly good info from dev mailing lists.
Yeah, Wayland is about as incomplete (in comparison to Xorg) as FreeBSD is (in comparison to Windows).
That's not true. IPv6 offers unique local addresses for building private networks. That's not exactly the same as the reserved IPv4 address ranges for private networks, e.g. all subnets usable with ULAs have the same size (/64, much larger than any IPv4 network) and the standard mandates using an additional random prefix in order to avoid the typical problem of having to "renumber" your whole network when two private networks are merged. Nevertheless, this is "the same with a few extras", so it can't be compared at all with X11 vs Wayland which is indeed a completely different protocol.A good comparable example would be the IPv4/IPv6 transition. IPv6 has no concept of 'private networks', unlike IPv4. Very different designs.
Maybe it's just me coming from from a long line of pure academics in my family, but there is such a thing as bias in research. One can come up with a ton of quotations from a ton of different places that support a pre-conceived idea (i.e. 'Wayland does not have feature parity with Xorg', 'Wayland breaks stuff'). A competent academic or dev can see through the nonsense and see stuff like cognitive bias and confirmation bias, which are very much present in the Wayland vs Xorg debates all over the Internet.that's why some of us posters (including myself and grahamperrin) quote information from developers who work with such software directly to further discourse; instead of compounding their own hypocrisy in criticizing other peoples take in shared interests. But of course, if it's not only from a dev mailing list (rather than a developers GitHub page); it's not a legitimate citation, right? right?
what?
Yeah, Wayland is about as incomplete (in comparison to Xorg) as FreeBSD is (in comparison to Windows).
Multi-faceted sarcasm did not register with the intended audience... There's may ways to understand this.what?
I didn't mention, and was not talking about, either of those but I see this all the time on low level boards similar to reddit. Such-and-such software needs to be replaced because it was created in the 1990s!! A couple of weeks ago someone complained cause some software was from 2009.old??? funny, graphics/blender (since 1994) is still 'new kid on the block', even though it's arguably just about as feature-complete as AutoCAD (since 1982). Nobody calls either of those two 'old' or 'crappy'.
Yep, and we do have debates about replacing ed(1), which is from 1973, or, more recently, about vi(1) (ca. 1976) (Thread is-vi-worth-learning-in-2022.87593)... those are older than I am...I didn't mention, and was not talking about, either of those but I see this all the time on low level boards similar to reddit. Such-and-such software needs to be replaced because it was created in the 1990s!! A couple of weeks ago someone complained cause some software was from 2009.
Think twice about Wayland. It breaks everything!
Think twice about Wayland. It breaks everything! GitHub Gist: instantly share code, notes, and snippets.gist.github.com
… many hours of reading enjoyment, for example, Think twice before abandoning Xorg. Wayland breaks everything! …
Maybe a better point of reference from the same author (I have not read either page, I'm in no rush):
- <https://gist.github.com/probonopd/d57e9c77e4fccc5e5f87d7000beb9730> – Wayland transition notes for X11 users
Exactly, you can find decent references from both side of the Xorg / Wayland argument. This is because, same as systemd, it isn't a clear improvement for many use-cases; it really is split.Maybe it's just me coming from from a long line of pure academics in my family, but there is such a thing as bias in research.
Kind of off topic. I recently was contracted to port the older Blender 2.49b to a current Linux and OpenGL 4.6 core (a client needed it for a specific rendering middleware). It took a couple of hundred patches. Agreed, it was feature complete back then, now they are just bloating it. I'm going to see if I can open-source it because I am sure many people would prefer this version.old??? funny, graphics/blender (since 1994) is still 'new kid on the block', even though it's arguably just about as feature-complete as AutoCAD (since 1982). Nobody calls either of those two 'old' or 'crappy'.
Disagree. For technical guys (who have the best chance to have a decent opinion about this stuff) they have been writing important notes in TODO files in VCS systems for decades. Doesn't make them any less relevant. Once it has bubbled its way outside the forges to "project management", then you know its usefulness is considerably reduced.Not an ideal point of reference. A GitHub gist was a poor choice of medium for something so controversial (and somewhat misleading).
? Current title, current heading? How is that adding something useful or any clarity to the discussion?For clarity:
Lina is definitely a different person than Hector. I believe the context of Hector's complaint is that he had recently been trying to get the Macbook Pro keyboard supported and the X keyboard system has like six bits or something absurd for scan codes and the Fn key has some crazy high number so making a driver was going to be literally impossible.The biggest example of this is that Asahi Linux developer (I believe Hector, Marcan or AsahiLina are his main aliases). What they wrote here about "unsuited to modern display hardware" is considerably untrue since Wayland compositors simply use Xorg's modesetting driver underneath anyway. Classic sleaze word "modern" used of course.
I believe (95.4%) Lina is a youtube streaming cartoon character (I think it is called a "vtuber"). It seems that marcan/hector is behind it. He doesn't seem to take advantage of it, but it would actually be funny to see him "self-cite" himself with regards to "how bad Xorg is" using these different characters / personas.Lina is definitely a different person than Hector.
I feel that is a much weaker reason. The way i.e VNC gets round this is a simple lookup table. Plus Xorg uses the same input system (libinput) as the majority of Wayland compositors.I believe the context of Hector's complaint is that he had recently been trying to get the Macbook Pro keyboard supported and the X keyboard system has like six bits or something absurd for scan codes and the Fn key has some crazy high number so making a driver was going to be literally impossible.
Basically compare the code between:I tried to figure out what you mean by "xorg's modesetting driver" but all I could find were wayland compositors claiming to use nVidia's X driver. Dunno if that really "counts"....
Considering the amount of work being done, I would not assume this.It seems that marcan/hector is behind it.
It wasn't about libinput, it was about an X protocol limitation after libinput.input system (libinput)