Hacker News Reader

Comment by phkahler | Hacker News Reader

>> Until Wayland has all these (and more) and they are as stable and feature-rich as the existing apps on X, I will not willingly switch to Wayland.

Every time this topic comes up, a bunch of people bring up the things X has that Wayland doesnt yet have, or have in the form they think it should.

I'm not sure if it's just me but these comments come across as some combination of defiant, demanding, or I dont know what. It's like they don't understand that the world is moving on and doesnt really give a $h!t about them and their love of X. I mean here is a post from "the guy" whose been keeping X going for a long time saying "this shit is done" and people keep complaining about that as if they have some say in it without actually contributing to the code. It reads like entitlement - some kind of expectation that other people exist just to make things the way they want them to be.

View on HN

The comment you're responding to lists very concrete issues. Try accepting constructive criticism instead of accusing people of entitlement just because they dare to disagree with you.

Entitled developers who think their work is above criticism just because it's open source are just as annoying as entitled users who think they are owed something.

They also seem to "dare" to disagree with major Linux distribution direction so perhaps it's not the responders opinion that's in the minority?

I mean, this really sounds very similar to systemd and pulse audio type gnashing of teeth where the new solutions added plenty of widely required features while compromising on niche use cases which had other workarounds.

So non laggy screensharing is niche use case? Such statement does not make any sense to someone that live in the reality, especially since covid.

Wayland bring the risk of having critical bugs on compositor Y and critical missfeatures/missoptimizations. Staying on X does not have any major user facing issue. Therefore choosing to use Wayland right now is an irrational action which pros does not offset the cons.

Maybe that it'll make sense in the next 5 years, then I'd reconsider

I've never seen non-laggy screensharing with X either (especially since most modern apps draw directly and don't use X commands) so that's a really facetious argument to make. Even with NX and Xpra.

State of the art has moved to image compression based screen sharing even on Linux and demanding that we stay on an obsolete stack which falls apart as soon as you connect a modern monitor to modern laptop due to some 1990s usecase is the crux of all these neckbeard complaints. Just like with Pulse, SystemD and bunch of others.

Linux distros want to stay competitive in the now, not in 1990s. Which is why they're gaining market share.

> Linux distros want to stay competitive in the now, not in 1990s. Which is why they're gaining market share.

This is where distro diversity is critical. Users have the choice to use a distro that is serious about maintaining backward compatibility or they can choose the ones chasing the new shiny. Everyone wins.

Personally, I’m not swayed by the “but it’s modern!” arguments. I just want my existing software to continue working. I don’t care if my stack is obsolete, whatever that means. I just don’t want to worry about software breaking when I update to the never version.

Other people are not like me and they can go use a distro with Wayland, and we are all happy.

I see major difference in there: both systemd and pulseaudio do things differently but in fact do not break any existing usecases. For both these "modern" solutions all the super niche things that get broken are still possible. While Wayland outright does not provide solutions for things that are decidedly non-niche.

PulseAudio certanly broke a lot of existing use-cases. But it also brought some very clear advantages.

Same with Wayland - it's a way to drag Linux desktop into a world where mixed DPI displays exist and where HW acceleration of desktop is a norm. That seems to be signifiantly more useful in 2020 where most displays need some kind of scaling.

>> Try accepting constructive criticism instead of accusing people of entitlement just because they dare to disagree with you.

That's valid, and I did say I'm not sure if it's just me - meaning my (possibly wrong) interpretation. A list of specific things missing is fine, but the I'm-not-switching-until sounds different than a simple valid criticism or an inquiry as to weather those are being addressed.

I probably should not have responded to a particular comment with my comment though. It wasn't too bad as those things go. Some of them really do come across the way I described.

I think this is a very emotional issue because people are being told that an important tool will be taken away from them, but I also think that it's important to think about being reasonable.

X, architecturally, is a dumpster fire. None of the ideas that were reified in it's design, types, or behaviour - not forcing rasters down unix pipes nor building widgets with X objects nor rendering networked fonts nor the general concept of immediate-mode widget rendering - are actually used today in modern X apps and have been mitigated-away with hacks and jury-rigs. But the design and the types and the protocol persists and no one wants to work on this nonsense in their free time. Unpaid programmers have voted with their commmits and Xorg has been orphaned. It's a dirty job and for the longest time only paid programmers have maintained X. And the last company to pay programmers to work on this stuff canvassed their customers and decided that the work didn't need to be done anymore.

Everyone using X today has been free-riding on RHEL customers for years and Redhat has decided not to burden them with this anymore. And nobody else is volunteering to pay for this.

I don't think entitled developers is the right term here. As an Xorg user, how much did you pay to use Xorg?

>Every time this topic comes up, a bunch of people bring up the things X has that Wayland doesnt yet have, or have in the form they think it should.

Every time this topic comes up, a bunch of people remind that Wayland doesn't yet have an accepted standard for basic features. Wayland is very good at some niche features like HiDPI or fractional scaling* , but more people (at least two orders of magnitude more) need screensharing these days when companies move to WFH.

* Yes, these are relatively niche features right now. e.g. Look at the Steam hardware survey: more people run resolutions below 1080p than above it - and that's in a biased sample, since gamers are more likely to have high resolution monitors! Now, it's nice to be future proof, but these features should not be prioritized over basic features.

Features drive adoption. I might be more inclined to believe in Wayland if other secured platforms adopted similar constraints. Unfortunately for Wayland, other comparable platforms have consistently found security-conscious solutions that don't compromise on added value.

Reply on news.ycombinator.com