💾 Archived View for gemlog.blue › users › BaronHK › 1692941092.gmi captured on 2023-09-08 at 16:18:32. Gemini links have been rewritten to link to archived content
-=-=-=-=-=-=-
Will Wayland Even Survive the Collapse of IBM? X11 Likely Will.
One of IBM’s stated reasons for pulling investments in actually important Linux software is they need to focus on Wayland.
I recently read on the mailing list, one of the Wayland developers basically admitting that they break the APIs and ABI all the time and usually don’t even get much “mileage”, so to speak, out of doing it. This shows, to me, that they rushed it out the door way before it was ready.
Developers tend to have noticed this. However, IBM doesn’t care. Their goal is to throw their weight around and make everyone use Wayland, even though it’s horribly broken.
All of the hideous things that X11 does were either fixed by extensions and patching over the last 39 years or don’t really matter much anymore.
In some cases, complaints about how X11 does things that “justified” starting over don’t even happen anymore.
For example, one of the Wayland developers (in an article that propagated to Phoronix) used an example of Flash and Java being subwindows of the Web browser.
Well, NPAPI (and Chrome’s version, PPAPI) is dead.
For better or worse, everything that an NPAPI (PPAPI) plug-in did is handled by the browser and the browser window now, so we no longer run into this problem in the window that IBM Red Hat says is the only one on your screen that matters.
(Because downloading an entire office program into your browser, which needs networking and someone else’s permission to continue running, is cool and modern.)
It dawned on me that while I was trying to get Fallout 4 working (in my recent post), almost all of the serious issues with the game that weren’t just some package that didn’t come pre-installed on openSUSE were actually Wayland problems, and not just Wayland problems, XWayland problems.
XWayland has its own special problems while running on Wayland that don’t happen when you just run your window manager on X11 directly.
Is X11 old? Sure. Does it matter much? Not if you ask me, as a user.
I would much rather have SeaMonkey and other X11 applications show up clean, crisp, clear, and scaled right, than smudgy and vaguely reminiscent of “Microsoft Glaucoma-vision”, which is what I refer to as the way Windows 10 and 11 scale things incorrectly and then put “Vaseline on the fonts”. This is just shoddy craftsmanship.
If you can believe this shit, Microsoft actually patented that (since expired), and called it a feature. It looked like the original edit of Star Wars where Luke Skywalker was hovering around on his speeder on Tatooine and George Lucas said they hid the wheels under it by smearing a blob of Vaseline on the camera lens.
Wine, which is a very important program to me, runs Windows programs. It uses X11.
When it runs on XWayland it picks up lag and stability problems and passes them along to the Windows programs and games I want to run.
When I double-click on Fallout or something, I want my game to run. Not skip frames and then jitter and die because XWayland doesn’t work properly, STILL.
One of the reasons I give to Windows users for why they should switch to Linux is that I’ve had really good luck with Wine. Wine is still not ported over to use Wayland and it’s not clear when or if it will ever be.
The openSUSE project avoided so much drama, so many difficult people, and so much political nonsense and technical problems by remaining with KDE and bothering to ship a version of it you can use.
One of the reasons IBM Red Hat dropped KDE entirely is that they were hoping to inflict a mortal wound, the same reason IBM supported defamation of an elderly man (Richard Stallman, who is 70.) and then stopped paying for GNU software development, but they still “freeload” off it.
Roy Schestowitz tells me he’s rarely seen IBM hiring Germans who have worked on KDE.
That’s fine by me if it’s true. That just means that we won’t have the most competent people going to work for IBM, being told to do dumb pointless redundant broken shit with GNOME and Wayland, and hiding under a desk when the Pointy Hair Boss goes by laying people off.
If Wayland still isn’t working right after 15 years, if it still behaves like some bugged, crummy, perpetual beta software, when will it work right?
I’m actually amused that IBM Red Hat considers it so production ready that it’s been in “Enterprise Linux” for a while.
Sure, as a user of an Enterprise Linux clone (or RHEL), you could fight your way like a salmon swimming upstream, using the last of its energy and time on Earth to lay eggs and die, fixing all of the really terrible engineering decisions IBM has made for you.
Some of the clones support BtrFS and DTrace (Oracle), and I think KDE is available as unofficial packages, but I think the Enterprise Desktop really deserves better software, and the developers of openSUSE realize that moving crackpot shit like Wayland over to a stable Linux distro just turns it into “Fedora with older packages”.
Some distributions, especially Red Hat, have this odd sort of definition of stability.
This definition means the software might be full of bugs, it might glitch, but at least it doesn’t change much. To those ends, IBM even ignores security patches.
And I guess I can see why you would shove Wayland in a desktop on an enterprise distro, when soon after, you divest from desktop work anyway.
It can load Firefox. It’s perfect. Firefox loads Office 365. Firefox and Office 365 are the only two things people should want to use.
What? You want to run software? On your computer? Oh…
It’s true that maybe in some tortured way, years from now, it will replace most of what X11 can do and, maybe someday Wine will work properly with XWayland. But it’s still not now.
Even on Fedora with GNOME I was still running into Wayland problems.
I wouldn’t be surprised if X11 outlives Fedora, RHEL, IBM, and on into the next century.
It’s become a fixture, like an old refrigerator that never breaks down.
KWin running so reliably on X11, mostly at parity with all of the redundant work just needed to get it working at all on Wayland (except that the Wayland version of the code isn’t quite stable), sort of proves that this has all been rather something of a misadventure of negative work that’s gone on for years.
IBM has been throwing out a lot of FUD about X11 being “abandoned”, but the mailing lists tell a story of something that still gets a lot of development attention considering that it’s 39 years old and feature complete, in ways that would only matter if you’re actually running it, and not just XWayland.
Even the Direct Rendering Infrastructure and DDX stuff for some video cards from the 90s just got updates last month. And that goes out pretty far from the core code.
As Free Software, X11 can live as long as it’s still useful. Nobody can “take it away” or force you to stop using it.
Finally…
Security! Everyone’s favorite refuge for when they’ve lost all their other arguments.
Sure Wayland can stop applications from reading input events from the other ones.
At least in theory, it can. Everything can be broken. There’s always bugs.
However, I just simply don’t care. This isn’t Windows (where applications can literally just dump dlls into each other and load malware into Firefox, for example).
I don’t have “Linux malware” because I haven’t installed any.
So to me, this “reason” to have Wayland is another fallacious argument.
This same “security feature” is one of the reasons why people have to fall back to X11 to make legitimate software work. So adding “security” that breaks too many real legitimate things is what Linus Torvalds (before they made him go to “therapy” he didn’t need”) called “Masturbating Monkeys”.
It’s up there with “Secure Boot” and “attestations”.
But that’s a whole different story.