Re: Wayland Is Coming

On Sun, 12 May 2024 08:48:49 -0000 (UTC), Muttley wrote:

On Sun, 12 May 2024 02:23:32 -0000 (UTC)
>
Lawrence D'Oliveiro <ldo@nz.invalid> wrote:

>
On Fri, 10 May 2024 08:34:15 -0000 (UTC), Muttley wrote:

>
If tha[’]s all it was it wouldn't still be in development after 15
years except for low level driver updates.

>
So do you think that “thin layer above the framebuffer level”, which
supposedly is “all that’s needed”, is still not ready for prime-time?

 
Its all wayland seems to do. Provide a low level buffer and let the
clients generate the pixels.
 

You think maybe it turned out to be harder to come up with than you
thought?

 
No, it turned out the Wayland design was a poorly thought out dogs
dinner and now they're trying to get it to work properly.

But you said yourself: “All that[’]s needed is a thin layer above the
framebuffer level to marshal graphics requests. Compositing isn't
complicated”. Are you changing your mind now?

You mean the X11 graphics API has been stagnant for decades. It’s just
useless baggage by this point.

 
Or possibly it[’]s stagnant because it works.

>
No, it’s stagnant because it’s been overtaken by other, far superior
graphics APIs. Like Display PostScript, OpenGL, Cairo, Vulkan ...

 
OpenGL was running on X back in the 90s FFS.

On top of, yes. As a replacement for, no. Given that you have something
that powerful, the X11 graphics API itself becomes ... superfluous.

Plenty of programs still use Xlib ...

Actually, a lot of them switched to XCB.

... and having the compositing side as an X extension ...

That’s been superseded by XWayland.

This is why, on Linux, you have a choice.

 
Until the zeolots decide X is for the bin just like they did init and
replaced it with the bug ridden POS called systemd.

Ah, seems you lack what the French call “Systéme D”. There is a discussion
about that in comp.os.linux.advocacy, if you want to go there and rant
about it.

Or we can rant about it here. I don’t really mind, though some others
might.