Magiilaro

There is a lot of valid points there yes, still don’t see where the old article would be relevant for those though, and I think that I really can’t say a lot against your arguments in general. All is clear now! Thank you for this longer clarifiation and have a nice day!

All I am saying is that his so ultra specific hardware works flawless under X11 and that a new solution that plans to replace the old one should support at least the same displays, that should be the MVP - don’t be worse then the old cruft.

And I still don’t see where the 8 years old article would say anything otherwise, at least nothing that was not replaced by info from the more recent article. And yes, he has linked his old article in the new one but that doesn’t mean that every single sentence in it is still fully valid after all those years.

I am not more ignoring it then you are ignoring the fact that the hardware article is 8 years old and that the current article makes part of the old one obsolete.

Well and his current articel from 8 years later says that it worked with X11 for years, so what? Maybe your arguments are a bit out of date? Information can be obsolete after nearly a decade.

Oh i read both articles and I still have no clue why it would invalidate my point. Please enlighten my stupid little mind with some proper context and not only a link to a long document.

Not sure what you mean but:

With the nVidia driver now working per se with Wayland, unfortunately that’s still not good enough to use Wayland in my setup: my Dell UP3218K monitor requires two DisplayPort 1.4 connections with MST (Multi Stream Transport) and TILE support. This combination worked just fine under X11 for the last 8+ years.

As he said, it worked in X11 but not in Wayland.

It is not wildly incompatible with almost everything, it works perfectly fine his current X11 setup, and any system or setup that has the goal to replace X11 should at least should support all physical displays that X11 supports. And a modern replacement should have no issues with modern hardware.

I am using timers for some time now, because systemd timers are included and I don’t have any good reason to install a cron also.

Maybe Canonical will expand into Hardware now and release a matching RISC-V computer too. Would match the “we want to be the Apple of the Linux World” vibe I get from Canonical for years.

He just turns into a Zombie Process and keeps going

I will never understand why it is a good idea to put SMB into the Kernel. Stuff like that belongs in user space, IMHO

There was a major focus shift maybe 5 or 10 years ago towards security in Linux design, especially with the development of Wayland, pipewire and systemd. The problem is that accessibility software behaves in many ways like spyware or malware. It reads all windows, it hooks themselves in programs, it redirects output and input. The security focused (even security first) approach of many developers broke all the accessibility workflows and proper API to do it the new and safe way have low priority. A few exist but it is still far away from feature parity.

That’s why I am against the Wayland default or even worse Wayland only approach that many distributions have nowadays, Wayland is still barely useable for many people who need working accessibility solutions and that should be seen as a major stopper issue for a wide release like that.