In general this sentence is why the Year of Desktop Linux won't come in this millennia. Not only XOrg vs Wayland. Many such cases. Sad!
But, Android is a thing, and Linux is literally everywhere. OS has won, even if the mythical "Desktop Linux" didn't.
After installing their first distro, most people have a good experience. Then they install an update, that causes some ridiculous regression. At this time they have 2 choices: spend 2 weeks on reading all kind of arcane documentation, or go back to whatever they were using before. Or ask the maintainer, who responds "deal with it".
Just as another, fresh example, look at this: https://blogs.gnome.org/mcatanzaro/2023/05/10/gnome-core-app... - Gnome finally has thumbnails in the file picker, but they removed the music player. You can't have everything, I guess.
This kind of attitude why the Year of Desktop Linux will not be seen by our generation, not evil manufacturers.
Do you have to burn a flashdrive image, navigate through odd commands to disable secure boot, and find the magical incantation to boot from USB to install a soundcard driver?
For example, look how saddled MS Windows is by the burden of binary reverse-compatibility. It makes the leap to other arches like ARM nearly impossible.
Yes, some features of X11 are not present in Wayland. But the incomplete reverse-compatibility allows us to decide if those features are worth it in the first place. And if you don't like it you can patch it yourself.
User experience is ALL that matters. Apple has this ingrained in their ethos. While the OSS communities are bickering over X vs Y, they chose to pick what's usable and polish it until it delivered great UX.
Last time I checked, Wayland was broken in fundamental ways on my NixOS/KDE machine. So I went back to Xorg, that for all its ancient faults, delivers a better UX _today_. I'll keep trying Wayland every few months, and when it becomes an improvement, I'll consider switching to it. No, I won't "deal with it". Why don't YOU fix it?*
*If you tell me that it's not up to you, but to application developers to add support for Wayland, I'll drop kick your sticker-infested laptop, and curse whoever came up with this ridiculous system.
I would love it to happen but the suggestion that there is anything the developer community can do is wrong. This isn't like the move to get friends and family to swap to Firefox in 2004, that was easy, it was in parallel with no data or app loss. With no new UX to learn.
I can get my parents to change to a different web browser. I could never get them to swap to Linux. I doubt I could get them to reinstall the os they currently use!
No you won't be able to edit system files on Mac OSX without rebooting into the bios effectively. Deal with it.
Every platform does dumb, annoying stuff. Deal with it.
Still, X.org is terrible and can be reasonably called the Internet Explorer 6 of the linux desktop. At least Wayland will be easier to displace if anyone comes up with something better. This is a step forward.
At work I have deprecated many older software components in favor of implementing newer, less painful to maintain ones. Lots of other software engineers use those components, so a big part of it is documentation or training. Rather than harp on the flaws of the old thing, I try to objectively present the lessons learned from it, with an emphasis on all the good aspects of the original that carried forward into the replacement. That generally makes the transition smoother, especially for getting buy-in from the folk that were involved with developing the original implementation. I also wouldn't tell someone not to use an existing component that works, unless I have a concrete recommendation for a better alternative.
And yet the average user will probably not do that, just like their eyes will glaze over if you say the incantation "soundcard driver" in their presence. It's unfamiliar and their friends don't use it, and that's it. Easy to install? Okay, but can you give a reason for them to even try?
Many people don't buy a Windows/Mac OS laptop or Android phone or iOS tablet. They buy a MacBook, Samsung S42, iPad, Steam Deck. OS? Web browser? Doesn't matter.
Do you have any open bug reports you can point to for "Wayland" not working? Which compositor were you try to use?
This all reminds me so much of ESM Vs cjs. Tech sometimes direly needs a kick in the ass. The situation cannot hold with something old & not good & the new works for most people, but also there's 1% valid problems. The world deserves to be able to eventually move on. The world can be bold & progressive & make changes, it should, and we should expect more of each other about being timely, about doing thr right thing, about making shifts. But we become so trapped by such resistance & unadaption, by staying where we are. Progress is a spirit I think should be Integral to open society & especially open source, but people don't see themselves as part of the bigger thing, they view themselves as merely at ends.
No. Why should a user go hunting for bug reports, or make an effort in making them to begin with? Developers shouldn't expect users to report every issue they run into. Most users will see something not working as expected, and move on to something that does. Fighting with software to make it work, a.k.a. "dealing with it", is a behavior exclusive to technical users and tinkerers. And even this cohort won't be bothered to do so when they just want their machines to work.
> Which compositor were you try to use?
How should I know? All I know is that I selected "Wayland" in my login screen, and ran into a half-dozen UI issues in less than 5 minutes, which I can't be bothered to list now (I posted about it here a few weeks ago if you're interested to know the details). Switching back to Xorg fixed all of them.
With the complexity of a DE like KDE, I, a technical user, wouldn't even know where to look for the culprit. Was it KWin, XWayland or Wayland? When Firefox had UI issues, is that because of Firefox, XWayland or Wayland? Or something with my combination of hardware and drivers? ¯\_(ツ)_/¯
Developers shouldn't expect users to debug their application issues. Users are not testers.
switching your desktop to Linux is still niche.
i think it being niche and also “the most popular it’s been” doesn’t have to be mutually exclusive.
That's great, but know that you're optimizing your QoL over the users'. If and when that new shiny thing becomes an objective improvement of the user experience, only then will users share your joy of switching to something easier to maintain. Just don't expect, or even worse, force users to switch just because it makes life easier for you.
> I also wouldn't tell someone not to use an existing component that works, unless I have a concrete recommendation for a better alternative.
Yes, this is key. And precisely what the person in TFA is failing to do. UX >>> DX, always.
- Chrome browser,
- Skype, Discord or other chat applications,
- update GPU drivers regularly,
- tolerate the abusive Windows update/reboot policy,
- install multiple anti-virus software just to fight with Windows Defender,
- etc...
but not willing to install a new OS, even if it's literally a few click.
I think we kept saying this "Linux can't be mainstream desktop player because it isn't default" so long, that we can't even imagine anymore that there can be other, stronger reasons too.
Defaults are definitely not meaningless, but users are willing to change bad defaults, if there is a better alternative. Sure, not your parents, neither mine. But most of people around you (I assume) and around me (I know) would.
Sure, just anecdata, but over time I have had more than 1 colleagues telling me proudly that they started to play around Linux on their machine, only to learn a few weeks later that they have abandoned it due to some (for them) insurmountable problem they faced.
You should know to be able to type `journalctl` which will almost certainly give you a report on what just happened.
Filing a big report feels like a pretty low barrier to me. It's stunning to me to see such a strong rebuttal, expecting developers to mind read & inuit every possible thing a user might run into. To shun openness, to remain closed & curt & expectant, feels more like a consumer relationship, where you don't have any real ability to speak or be present. But do have that ability, you can provide info that would make things better for everyone, and to me that barrier to helping feels extremely low. Not a burden.
That’s FUD. The reason around the controversy about screenshots is as simple as gnome wanting to do it right from first principles, that is to also include proper support for taking video recordings/streaming, which is much more complex, while wlroots were looking for a simple “take a picture” mode at first. They couldn’t come to agreement on a common protocol for a time, end of story.
This is due to the bazaar style of development, has no technical reason. Apple can just say they will switch to this new tech in x time, linux can’t really.
(these are not necessarily the same issues, but at least similar-sounding ones I found by googling)
• widgets opening in wrong place: https://bugs.kde.org/show_bug.cgi?id=459188
• ignoring icon size configurations: https://gitlab.freedesktop.org/wayland/wayland-protocols/-/i..., closed as wontfix
• on touchscreen menus open wherever you left the cursor: https://bugs.kde.org/show_bug.cgi?id=406025, more info in https://github.com/PeterCxy/evdev-right-click-emulation/issu...
• cursor framerate issues: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/1827 maybe, it's in the gnome issue tracker but described as wayland-specific, so might also be happening on KDE
• firefox using wrong screen's DPI: https://bugzilla.mozilla.org/show_bug.cgi?id=1632829, resolved by original reporter but last comment is a repro from someone else. also reported in https://github.com/flatpak/flatpak/issues/5300 but closed as "wrong bug tracker"
• fullscreening a video rotates it from portrait to landscape: https://bugzilla.mozilla.org/show_bug.cgi?id=1600962 possibly, https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/190522... possibly
In general wayland seems too buggy for major software like firefox to use it by default, per https://askubuntu.com/questions/1409389/why-doesnt-firefox-r...
Impressive list to hit in only 5 minutes!
Some of those might be the bugs I experienced, but then again, some of them claim to be fixed years ago, and I was running a fully up-to-date NixOS unstable (rolling release) system.
Reporting an issue in an OSS project is a great freedom that users have, but it involves so many steps, especially in these complex environments, that it's often not worth the effort. Even if you manage to figure out which component is misbehaving, to search if there's an already open bug report, and take the time to create one if not, there's a chance of running into hostility and WONTFIX responses.
> Impressive list to hit in only 5 minutes!
All it took was logging in, launching Firefox and playing a video. If these issues were so obvious in 5 minutes, I couldn't imagine how many there would be had I used the system for longer. Switching back to Xorg was the quickest fix to my workflow. Besides, I figured that since these issues were so obvious that they were already known. I was just befuddled as to why they remain so prominent given that Wayland is already 15 years old, and there are mainstream distros offering it as an alternative, or switching altogether to it.
We're probably never going to be in a situation where someone can "implement Wayland" and have an acceptable desktop. The real standard to be implemented was written as part of the sway team circa 2018 (eg, https://gitlab.freedesktop.org/wlroots/wlr-protocols/-/commi...).
Of course if the user is technical enough they can figure out how to resize their existing partitions & set up a dual boot. But that's not the default, and that still risks corrupting the existing partition and therefore deleting all their programs & data.
I feel like people spend more effort trashing & talking down software in forums like this, than it would have taken to make an even mostly-constructive tiny little data point that could actually help open source, and open society. The author, a NixOS user, almost certainly had plenty of basic technical chops they could have brought to bare within mere seconds to get some first level idea what happened, which they refused. The malice/concentrated-negativity, combined with the lack of effort, to me, invalidate my willingness to take someone seriously.
There are plenty of good valid useful ways to share the view & be part of open society. Just trying to tear out the supports of the building & make it crash down- anger- is not an acceptable enduring view; it's an ok initial reaction, but eventually realer more mature cycles have to begin (unless despair & ruin is the objective)
As a passionate long time supporter, user and developer of OSS, my reaction was purely based on that "deal with it" quote from TFA.
All software, regardless of its licensing terms, should first and foremost deliver a usable product. The expectation that users of free software (in both meanings of the word "free") are somehow under moral obligation to test the software and report issues to the authors is wrong, just as OSS developers aren't under any obligation to implement specific features or provide support. These freedoms go both ways.
But when a developer takes the hostile stance of prioritizing their QoL over their users', as in TFA, that is the attitude that's hurting users, and the adoption of OSS. I appreciate the effort the Asahi team is making, but this is a toxic attitude that shouldn't be part of any project.
Most of the time, I do take the effort to diagnose issues and report them where appropriate. But as you can see detailed below[1], the issues I experienced were so obvious that I assumed they had already been reported. Given the complexity of an environment like KDE, it would have definitely taken me hours of my time to properly diagnose and report these issues. And sometimes, I just want to get work done, without "dealing with" my machine... The reality is that most users feel this way, so this response shouldn't be surprising.
User: I tried it and it didn't work well don't care whose fault it is I just care that it didn't work.
Response: Every developer that flings non-working buggy software into the void is entitled to the free labor of their users who will test said software and write bug reports so that in the fullness of time it may someday be worth using. If you by the mere act of downloading something once don't see yourself as part of that developers community then your opinion is clearly invalid because you are like unto a leper unclean and outside the nourishing light of our lord!
A proper reproduction of the bug, troubleshooting, research into existing bug reports and possible causes both in the immediate and upstream project and writing it up could easily be half an hour not seconds. Meanwhile 17 other things have claims on our time. It is perfectly acceptable to shoot the shit on a forum about how you dropped a non-working project like a chicken nugget with a human finger in it without also opening an issue.
Do you know what people do when they install your software and it crashes on first run. They delete it without comment because they are not indebted to you by being granted to privilege of non-working software and running something else is liable to be a better investment of their time. This isn't abnormal or malicious its absolutely normal. This is also why your software which has a bug that effects 10,000 people gets 5 bug reports. It's not because the other 9,995 people noticed the bug had already been filed its because none of them bothered.
Pretending this is something abhorrent and invalidating is quite frankly a gross failure of analysis committed in service of invalidating a negative perspective that seemed to stick in your craw.
But quickly their position crumbles. No advocates for their position maintain. What they were doing stops working. The user is swept under the sands of time & forgotten.
This putting all the onus on the developers is insufficient. The user doesn't have to advocate for themselves, no. Time happily can ignore their plight. We don't need them. The rest of the open society world is participating onwards.