[ethin]

And that is exactly the problem. Now things like accessibility (or, really, any feature that the maintainers of the core protocol didn't think were "necessary" because they like minimalism) have to be implemented by each and every compositor. If the compositor doesn't implement it, your screwed unless either you convince them to add it or you add it yourself. Talk about causing huge amounts of fragmentation for absolutely no reason.

The proper thing Wayland should've done is waited until Wayland had reached feature parity with X, then released it to the world and started acting like it's the future.

[kmeisthax]

Wayland was specifically built to support things that aren't desktops, so feature parity with X was never a design goal of Wayland. The idea was that Wayland would be a super-flexible "you give me a window and events, I give you rendered bitmaps" kind of protocol, and then desktop functionality would be layered on top for people who wanted a desktop. Not everything needs to be a desktop (e.g. car infotainment displays, KDE Plasma widgets, etc), and some protocols would be super limited if they had to fit in a desktop mold (e.g. VR displays[0] with apps in non-planar windows).

The main mistake FD.o made is that they didn't get consensus on a "Desktop Profile" extension, so all the DEs wound up implementing their own thing. This is still fixable, just very annoying until we have agreement on this shit. I think that's what you meant by "feature parity with X".

[0] Currently, every desktop VR setup has to have two layers of compositors. VR applications have to communicate with a special VR compositor that then draws normal desktop windows with the contents of what should be hitting each eye of the VR display, all so it can pretend to be two normal displays.

[somat]

I could say the exact same thing about X, A lot of the problems people had with X historically was that developments goal was to "create mechanisms not policy" and people just wanted a desktop environment that worked.

An antidote on non desktop use of X: the other day I wanted to show a program on my phone, there are many good ways to do this, but I picked none of them. Instead I had just installed a terminal on the phone and noticed they had an X11 package, So A few minutes later I was the proud owner of an X server on a phone. And you know what... It was pretty great. My gaming system load and temps dashboard were displaying just fine.

Despite using X for many, many years, I had never just sat down and played with a bare X server, I had only dealt with it through the lens of a locked down, encumbered desktop system. It was like having a network attached monitor. From whatever system I was using as a desktop system I could just go "display this on that monitor", in this case a phone. Based on that experience I put a raspberry pi on my TV running a bare unprotected X server because having a network attached monitor rocks.

[ethin]

Yeah, pretty much. I would be less disagreeable about Wayland if they had solved this problem early (and yes, they should've thought about this early during Wayland because the most prominent target is desktop environments). But they didn't, and I don't even know if they'll come up with some unified solution that all DEs/WMs can agree on or whether they'll just keep allowing DEs/WMs to do their own thing. Either way, fragmentation is never a good idea on what, I think, many would consider critical functionality. At least, I consider the requirements to implement accessibility to be rather critical, which is the primary reason I still use Xorg.

[gf000]

If I have a kiosk terminal, why would I want the overhead of, say, screensharing? Also, isn't this the point of libraries, so that you only have to implement stuff once, and you can reuse it in different projects? Like you can build on top of wlroots just fine.

> The proper thing Wayland should've done is waited until Wayland had reached feature parity with X

How on Earth would you expect a fundamental protocol to be developed behind closed doors?! Wtf even.

[ethin]

> If I have a kiosk terminal, why would I want the overhead of, say, screensharing? Also, isn't this the point of libraries, so that you only have to implement stuff once, and you can reuse it in different projects? Like you can build on top of wlroots just fine.

Yeah but again this fragments the ecosystem massively. If people really wanted flexibility they could've just made it a configure option or something equivalent?

> How on Earth would you expect a fundamental protocol to be developed behind closed doors?! Wtf even.

Your making a pretty big assumption here, aren't you? I never said it had to be developed behind closed doors. It's the "lets just obsolete X11 even though Wayland can't even replicate a quarter of it's functionality right now now now because of security security security" that irritates me. If they had worked on Wayland and obsoleted it once they had reached feature parity, that would be releasing it to the world. Then they would've had far less friction and the transition would've been a lot smoother. Would it have delayed Wayland by maybe a decade? Sure, but I see little issue with that. IMO that probably would've made Wayland even better.

[gf000]

Who is the supposed agent mastermind singlehandedly developing Wayland and deprecating X11? You do realize there are multiple people working (having worked) on both in their free time and each have agency and their own incentives. Separate people have had enough of maintaining X, while another group of people enjoyed working on Wayland. Some indeed moved from one to another but there were no coordinated attempt at hijacking the Linux graphic stack or whatever..

[kasabali]

So you tell us Freedesktop.org people

1. Claiming XFree86 evil

2. Forking it as X.org

3. Shortly after all distros finished switching to X.org, declaring it obsolete and announcing wayland

4. stopping any major development on X.org immediately even though it's was the one and only option at the time

5. and channeling all development resources (not only on the display server, but also downstream users like toolkits, DEs etc.) to rewrite their code for a protocol that wasn't even gonna be usable until a 10+ years later

6. Depraving Linux desktop users from 10-15 years of improvements and making Linux graphis stack stuck in 2000s

wasn't hijacking the Linux graphic stack?

I mean had Steve Balmer wanted to sabotage Linux in desktop he couldn't do better

[tuna74]

If you want to continue working on Xorg you can. You could have done that 10 years ago as well.