Live data from Hacker News

I'm tired of this anti-Wayland horseshit

drewdevault.com

261–270 of 892 posts

Re: I'm tired of this anti-Wayland horseshit

#261

Wayland is being treated like PulseAudio and systemd before. So, it will probably win its battle, thanks $DEITY.

That's because all three are similar, all three throw away some of the modular unix philosophy and replace it with a system that gives less power to the user.

The modular linux philosophy is great, but it is quite obviously not the optimal solution to all problems. Some problems are better solved in a non-modular way.

Re: I'm tired of this anti-Wayland horseshit

#262

Earlier quoted context omitted.

And that punting is indicative of the problem. The reason the stacks we have are hot garbage is because they didn't have the luxury of punting on hard problems, and hard problems solved quickly make your stack hot garbage. It does give you a garbage stack that solves problems though.

I don't see how screen shots/sharing is really that hard of a problem. Wayland's model doesn't give the ability to arbitrarily capture these things the way X did, but surely it couldn't have been that difficult to add an escape hatch for common use cases like this to the base protocol. Remote desktop is a harder problem to be sure, but it is also a pretty common use case (window remoting less so). If you're setting o…

I'm not a developer on Wayland and I haven't looked at the code, so I can't comment directly on their situation. A hypothetical I can come up with is that if they've tried to implement a very secure architecture, they may have painted themselves into a corner if there's no easy way to capture pixels post-composite... Then no one piece of the system has enough authority to see all the pixels to capture an arbitrary screenshot, and oops.

An unsecured screenshot escape hatch can be used by malicious software to scrape information shown in windows on your system, and with modern OCR being what it is (or, hell, a shady Mechanical Turk transcription job), that could be an attack vector for surfacing people's private information.

Re: I'm tired of this anti-Wayland horseshit

#263

Earlier quoted context omitted.

> I can certainly understand the frustration at having to deal with a community that gets annoyed at anything that changes simply because they're already so used to the garbage pile they have that they now hold the delusion that it's actually not garbage. It's not just the FOSS community, look at how non technical people react every time gmail, outlook or excel makes a tiny change to the UI (not to mention big ones).…

> It's not just the FOSS community, look at how non technical people react every time gmail, outlook or excel makes a tiny change to the UI (not to mention big ones). It really makes people angry when things don't work the same anymore, let alone when it doesn't even work anymore for their use case. Like you note, this isn't a problem unique to Linux/"free Unixy OSes". In some ways Linux is actually worse than others…

I think a big issue is there's no discipline in the Linux community. The Kernel team has done an amazing job of preserving ABI compatibility through every last byte, but then you look at projects like glib or GNOME which seem to pride themselves on huge, sweeping, changes and don't care if downstream consumers are hurt by the changes.

Edit: glib not glibc

Re: I'm tired of this anti-Wayland horseshit

#264

I think Wayland has come too close to the following pattern, which I've seen in other cases of free software which have become controversial: - decide it's time to make a New Thing which will supersede the Old Thing rather than living beside it - decide that some features of Old Thing will not be directly supported by New Thing; instead New Thing will be flexible enough that people could implement those features on t…

This is exactly the kind of post that the author was talking about.

"decide it's time to make a New Thing which will supersede the Old Thing rather than living beside it"

A non-substantial argument. Whats the difference between a wayland 'superceding' X and a wayland 'living besides' X? Does implementing Xwayland not constitute trying to 'live beside' it? your argument is vague and you can respond with more ambiguity. Install X and wayland. Nothing stops you.

"decide that some features of Old Thing will not be directly supported by New Thing; instead New Thing will be flexible enough that people could implement those features on top of it"

WHICH features? the 'old thing' can never equal the 'new thing' otherwise they would be the same. Wayland is purposefully LESS flexible (again, how is wayland more flexible? Or is that another vague statement?) in its input handling and screen handling which is why screen recorders needed to be reimplemented in the name of security.

"declare that the work done to support those features will have to be done by the people who want to use them" How is this controversial? You seem to ascribe some kind of motivated reasoning for the statement that certain features aren't being maintained/developed by wayland developers. Dev says 'no' to a feature request, why is this controversial? EGLstreams is the closest thing to a controversy but nothing prevents anyone from forking and implementing support. Open source, no excuses.

"declare that the time has come for New Thing to supersede Old Thing, before said work has been done"

no such declaration has been made.

Re: I'm tired of this anti-Wayland horseshit

#265

Earlier quoted context omitted.

I don't understand the point about maintenance. Is Xorg maintained? If so, are the current maintainers handcuffed to it? I assume they aren't, so what's the problem? If Xorg is not maintained, again, what's the problem? Software is known to outlive its creators and maintainers. If there's a need, it will eventually get new maintainers. Software is not magic.

X is barely maintained anymore. To a large extent all of the X developers are working on Wayland now. A lot of people have the impression that Wayland is some outside competitor to X but the truth is that the vast majority of the people who started Wayland are long-time X developers who know its flaws inside and out.

Sure, I can understand people working on software X for a long time getting fed up with it and start working on software W.

I can also understand that these people have strong opinions on why W is so much better. The only trouble is that others, who are possibly not developers of window systems, may not care, or may find that it breaks their software, or may not want to write software that depends on W and doesn't work on X.

So what if Xorg is "barely maintained"? It works. Hell, as long as it works, it's likely better for it not to change. If there are issues, in the free software world we say "patches welcome". No developer should feel chained to a project.

Re: I'm tired of this anti-Wayland horseshit

#266

> In the case of Wayland, the “vague authority” are a bunch of volunteers who have devoted tens of thousands of hours of their free time towards making free shit for you. That does not mean anyone is under an obligation to like it. > Maybe Wayland doesn’t work for your precious use-case. More likely, it does work, and you swallowed some propaganda based on an assumption which might have been correct 7 years ago. If y…

> "well you don't need that anyway!" Something similar used to frustrate me no end when asking the Python community for help trying to solve a problem. Regularly someone would comment, "Why would you want to do that?". I wouldn't be asking if I didn't have a need to do it! It's bizarre that some people can't see beyond the end of their own nose and imagine different use cases to their own.

But, it's also a reasonable response to prevent bloat. Not asking that question is how anything with backwards compatibility guarantees, especially languages, get too complex over time. Asking the question doesn't mean you've hit a wall; the maintainers may just be going through a mental checklist to try to keep this from being a major pain for them in 10 years.

Re: I'm tired of this anti-Wayland horseshit

#267

> In the case of Wayland, the “vague authority” are a bunch of volunteers who have devoted tens of thousands of hours of their free time towards making free shit for you. That does not mean anyone is under an obligation to like it. > Maybe Wayland doesn’t work for your precious use-case. More likely, it does work, and you swallowed some propaganda based on an assumption which might have been correct 7 years ago. If y…

> "well you don't need that anyway!" Something similar used to frustrate me no end when asking the Python community for help trying to solve a problem. Regularly someone would comment, "Why would you want to do that?". I wouldn't be asking if I didn't have a need to do it! It's bizarre that some people can't see beyond the end of their own nose and imagine different use cases to their own.

The people replying like that are in the barely-not-novice-anymore category themselves, and their mental model of the world is that everybody is just fumbling around doing everything wrong like they were doing themselves a short time ago. It's true that when you are just past the point where you're not totally useless any more, there is a point where you think you're God's gift to the world (not judging - I have felt like this many times over, it goes away after the next big hurdle you run into...), and it's impossible at that stage to imagine the type of abstract thinking you get after many more years of experience.

But yeah - the 'XY problem!' crowd (along with the 'But The Rules!' crowd) has also totally destroyed Stack Overflow.

Re: I'm tired of this anti-Wayland horseshit

#268

Wayland is being treated like PulseAudio and systemd before. So, it will probably win its battle, thanks $DEITY.

I still have to kill pulseaudio regularly to get audio working. In 2021.

I'm pretty sure you can automate killing PA with systemd.

Re: I'm tired of this anti-Wayland horseshit

#269
post #14

I’m using a Dell UP2718Q (3840x2160, 144dpi, 60Hz) and a Fujitsu Siemens P17-2 (1024x1280, 96dpi, 75Hz) side by side, and without Wayland this setup isn’t even possible to get working properly. Sadly, KDE still doesn’t support it nicely under Wayland (so I moved to Ubuntu Budgie), but under X11 it’s entirely impossible to run it properly, as except for Qt nothing supports mixed DPI, or mixed refresh rates properly. O…

I get the frustration, but fundamentally, screen sharing and screenshots are so integral to my daily use that I couldn't even consider wayland. For nearly any remote worker at this point, at least the ones I work with, this is fundamentally true. In 2021, screen sharing and screenshots are table stakes.

I use KDE spectacle under Gnome/Budgie for screenshots (works fine) and screen sharing works fine with Slack, Teams, Discord and Jitsi.
Post reply on HN