>valve dev reverts wayland preference over X11 on the next libSDL release because IT'S STILL a broken mess of software unless a feature is merged
>feature is still to be merged since 6 months
>they're discussing english terminology on GitLab for over a month now
its so over for wayland cucks, Xorg won
 Thalidomide Vintage Ad Shirt $22.14 |
 Tip Your Landlord Shirt $21.68 |
Thalidomide Vintage Ad Shirt $22.14 |
Time to use gamescope on my desktop.
Gamescope is really nice for running old windows games with wine.
Game crashes and your desktop stuck 320x240? Gamescope doesn't care.
Game is windowed only? (Or fullscreen only).
homie, I control what is windowed and what isn't.
Game is 4:3 and gets all stretchy when fullscrewn? Gamescope scales shit properly.
Can I run touhou on it?
Yes
2hu was the reason I first tried gamescope because those games wouldn't run fullscreen without stretching otherwise.
Tried it but it has horrible screen tearing problems.
try it on a dedicated tty. its mind blowing how much a window manager, x or wayland, affects latency.
Latency is theoretically non-existent but some compositors are shitter than others.
I expect you'd get more performance not from the fact that it's running in the TTY but simply because you don't have any of the other desktop services running anymore. No shell, panel, background daemons, file indexer, etc, just Gamescope.
this is why I run xorg without a compositor, I cant go back
>dev politely explains the issue and why they have no other option
>30 downvotes
mental illness
It's called reddit mental illness because i don't like what i'm reading so i must downvote so express my disagreement
They don't even engage with this in the comment section.
That's the thing. They express disagreement without saying why, even when the message from that dev is crystal clear, explains the issues there's is currently with Wayland in an informative way and also explains the steps needed in order to fix it. But boo hoo it doesn't say what i want to read dev bad.
Maybe they just don't feel like repeating what's already been said. Like the fact that this doesn't affect OpenGL at all and only affects Vulkan, etc.
It is a real issue but whether or not it'll affect you depends.
Incorrect. EGL is affected as well.
Then why are people in the issue saying the opposite?
The issue as far as I understand it is that the Wayland WSI fundamentally disagreed with the Vulkan spec. This is a problem for anyone use Vulkan with FIFO but not with mailbox.
>Then why are people in the issue saying the opposite?
Nobody said that. Vulkan just happens to guarantee forward progress so it actually violates the spec. EGL technically doesn't but it doesn't mean that it doesn't have this exact same problem too with swap interval 1. It does and it's exactly the same issue. The frame will be blocked indefinitely if the window is occluded.
>the frame
thread* to be more correct
That's the issue of votes. It provides a way for people to have an opinion without having to put it into words or argue it.
It's basically 37 replies that said “^this” and 30 that said “no u”. That's the kind of people a voting system attracts.
wayland troons are a cult of clueless morons
Most of the time they're also amdrones if you check places like moronix.
Wayland is still broken shit on amd. The amount of weird glitchy crap that happens is baffling for something being shilled so hard.
why you have to bring troons into this? what'd they do?
Fricking women, for god's sake...
That's a man
it's like I'm back on facepunch
In retrospect it's completely unsurprising that it turned into the complete shit pile that is knockout.
Why does Reddit and Github even have a downvote?
YouTube is much better now without the negativity.
pussy
IQfy doesn't have a downvote BTW :^)
It's not a downvote on github it's an emoji reaction like on Discord
>There is no advantage to games and average applications preferring Wayland over X11 -- only severe performance and unusability regressions right now.
He's 100% right.
God he's so based.
>NO! You must force people to use broken crap to signal that actually it's really good and we'll fix everything at some point
Why are Waylandgays like this?
It got them this far
took notes from systemd
Probably took funding from the same place as systemd as well.
>Based for saying that software should be for it's consumers rather than being ideological.
That's just being normal. I think you interacted too much with Freedesktop and your standards got low now.
believe it or not the majority of programers use to think like this a decade ago now its all forced ideological bs and screw the user
while I admit it was better a decade ago, the majority of devs were still dumb trend followers then too. Generally speaking it's a sign that the developer is good if he has idiosyncratic opinions.
It's always been this way in free software.
It's been like what, three decades ago now that GCC refused to have a plugin interface out of fear that it could be used to be benefit of proprietary competitors?
what the frick do you want a plugin interface to GCC for lmao
You can look at how Microsoft/Nvidia/Amd/Intel all have compilers base on their LLVM frok
It was mostly about Emacs and other text editors.
Essentially the plan was to integrate them with GCC and use parts of its code for better syntax highlighting and source code manipulation and in-editor type checking but it was abandoned because Stallman vetoed it as the plugin system could be used by proprietary competitors as the GPL wouldn't be contagious for something like that.
Were you ever born a decade ago? Not caring about user is a long tradition in software development. Even more when it's freetards
>customers
That is were he is wrong and why Wayland is the way it is. Freetards don't understand that without money there is no incentive to cater to the users, there is zero reason to work with or for users, because guess what they are paying you nothing for your work and their requests. The only motivator for open source autism wagies is that they can impose their own ideology, vision whatever on the project, and on the people using it which are in turn again freeloading neets. If people were paying for Wayland then this shit would be solved in a few weeks.
Meanwhile, in the real worlds it's Redhat pushing this shit onto paying customers.
Redhat makes money from providing support.
Weirdly enough, it has an incentive to produce a broken product.
The weird thing is that many of the hobbyist projects by people done in their free time actually do implement features rapidly because they're not commercial and have no ideology.
Someone uses a some free software and is like:
>Wow, this is really cool, but I wish it had this feature too.
>I know, I can just write it...
>Wow, it works, and makes it so much better
>Others should have it too, let's send a pullreq
>Wow, this is a cool extra feature you added in. I don't personally use it but what's the harm in merging it
>boom
Whereas with Freedesktop
>NOOOO, NO EXTRA FEATURES. TOO MANY FEATURES CONFUSE OUR USERS, NOOO
>NOOO, NOT TOO MUCH CUSTOMIZATION. WE WILL LOSE OUR BRAND IDENTITY THEN
More features = more buttons
More buttons = less padding
less padding = less modern UI
That's how UI/UX works now.
Most of these features are just a line in a config file anyway.
But easily editable plain text config files are “confusing” or something.
Nah all settings are supposedly be available on some UI, in a setting panel with as few buttons as they can get with or even not having settings cause there's no features at all.
That's the gnome way.
Your entire scenario relies on there being another autist who works on the weekend on his pull request because it is a feature he wants, and that there is the owner who accepts the request into the project. A user who does not have the time or skill to work on a feature will be left stranded until that happens or even worse it never happens because the pull request is rejected or ignored. Open source as a model relies as an incentive on being able to impose whatever the owner wants or wishes, there are no monetary constraints or incentives which makes it less likely for quality software to be developed and even less likely to develop for what the actual users want. Note how all quality open source projects ultimately require a dedicated financial backing to fund people to work on the project and the features users actually use. Blender only is the way it is because it has income which it can use to hire developers, same for Firefox, Android and so on.
Well in practice it works. It's not a model or scenario; it's an observation.
In practice all these hobbyist projects pretty much accept any feature anyone adds to it except for things by Suckless which again is more politics than anything else.
All those commercial Red Hat isms often reject entirely useful features someone has already made over political or commercial concerns such as “It will harm our brand identity if people can configure this”.
And yes, someone has to write it but statistically, if something is in demand enough then someone with the skill to do it will show up
>Note how all quality open source projects ultimately require a dedicated financial backing to fund people to work on the project and the features users actually use. Blender only is the way it is because it has income which it can use to hire developers, same for Firefox, Android and so on.
That's simply not true. In fact, when Wayland first came out I feared it would be impossible to implement compositors without such funding but that hasn't been the case and I severely underestimated the fact that of course some kind of library would emerge like wlroots which would help it.
Even Wayland compositors are made by amateurs in their free time without funding.
This may interest you:
https://github.com/winft/como/
additional link:
https://www.phoronix.com/review/the-compositor-modules-como
>Your entire scenario relies on there being another autist who works on the weekend on his pull request because it is a feature he wants, and that there is the owner who accepts the request into the project
As someone that develops for a moderately popular FOSS project, this is basically what exactly happens.
>muh pwofit motive
Meanwhile in reality, Wayland's development is funded by IBM/Red Hat, along with other abominations like GNOME.
>he knows
>not blocking github.com##.js-reactions-container
You need the peanut gallery to tell you how to react, you fricking gays?
>troony shark pfp
Opinion discarded
he's not actually a troony
>not your tool for "signaling to stakeholders" about what is important
Can you do that? Just toss the current societal paradigm in the trash? What will happen to Our Democracy?
Can't wait to piss in the face of that snide aussie waylandshill
>stakeholders
Holding stakes in what exactly? Valve is a private company
stakeholder =/= shareholder, my Black person
The stakeholders are the company itself, its employees, its customers, developers using its software, the customers of companies that use its software, etc. as Valve publishing bad software will badly effect all of the above.
Lad should grab Ian Sommerville's tome.
Don't worry I'm sure it will only take a few more years of sitting on merge requests to reach feature parity with X11.
so basically don't use any linux distro which uses wayland for gaming, got it
do NOT redeem the pr saaar
pleas gib wayland a chance sar it the BEST THING that happened to linux
i know SDL3 its still not production ready but DO NOT REDEEM THE PR SAAAR
Kek
Does anyone have a screenshot of the guy who made a pull request that sent hundreds of thousands of emails and broke github for a few hours
all for his "gorgeous" readme. What a hero
> just destroy your library for the wayland cause!!!
sounds like the endless systemd propaganda everyone fell for. "hey systemd breaks our library so we aren't going to use it." "no you don't understand how to write a library just break all your apis to work around us changing how the linux user environment functions on a fundamental level"
>muh sysvinit
Lack of a feature doesnt mean broken, you fricking troony, im tired of you people bei g hysterical about this shit
It is broken though.
Those extra protocols are just hacks to work around how fundamentally broken Wayland is.
Too bad there's no protocol for mentally ill trannies like (You)
That's called a gun-to-your-head-v1 protocol, or if you go with gnometards mentality, a good-old-rope-v1 since you dont need anything extra.
They were hacks in x11, everything is a hack
Xorg was full of hacks because X11 is a legacy protocol, gayland has no such excuse.
Welcome to reality.
Frick off, "reality" is not forcing the Wayland maintainers to be the most intentionally difficult people alive. Literally every major Wayland compositor has ad hoc extensions to the protocol suite, purely out of necessity, and the response to this is always some short-sighted bullshit like "Wayland is just a protocol" or "just use dbus lol".
That is the reality. You can be an agent of change or cry and cry like most Americans when reality spits on their face.
>You can be an agent of change
By doing what? Finding where they live and holding them at gunpoint?
That's one option.
>X11 is a legacy protocol with hacks
>Wayland is a modern protocol with hacks
Wew
It's such a fricking mess at this point. To be wayland compliant compositors and toolkits need to support
>CSD and server side decorations
>implicit sync and explicit sync
>integer scaling and arbitrary scaling
At least X11 legacy cruft has a use case. The stuff making wayland needlessly complex to implement is all because of bad, easily avoidable design decisions made over a decade ago.
>bad, easily avoidable design decisions made over a decade ago
sounds like legacy cruft to me
wayland is over 15 years old, its legacy too
We really need to get rid of all this legacy spaghetti code and switch to Arcan.
Stop trying to make Arcan happen because it is not going to happen.
I'm writing a wm for it tbh it's pretty comfy.
Post repo.
Sorry but no. I am not the kind of person that collaborates or even cares to have other people look at my personal projects, I'll have it up when I consider it 1.0 and you can fork it if you want to change it.
Yeah and instead of making wayland with those hacks in mind, it has now has even bigger set of shittier hacks
Never understood how mentally unstable wayland schizos were until this PR. Fricking hell, man
https://github.com/libsdl-org/SDL/commit/254fcc90eb22bb159ab365ad956222a9c5632841
>merged
It's over sisters.
wonder what GNOME's response will be.
what's the use case for a response
They need to ask 20 different cripples for their opinion because as they've shown in the past they don't give a shit unless a cripple is affected by something
That's from 2022, sister.
>wayland
leave it to open source to frick even greenfield projects up
>leave it to
>open source
redhat* footgays* trannies* communists*
and a lot of people getting paid for this shit that will be out a job if wayland ever becomes usable and stable to develop for. let's watch and see how many of them embrace kde's windowing library, i bet they all reject it in the name of some backwards principle, but really, because it just works
Why the frick is it taking Wayland so god damned long to be good. I'm sick of hearing about this pile of shit and all the problems its causing.
Is it another one of those poorly thought out hipster projects where they left out some glaring requirements and tried to reinvent the wheel?
much like the GNOME Project, most of its contributors are "volunteers" who receive incentives from red hat to disrupt the desktop Linux ecosystem
>where they left out some glaring requirements
its basically this
they're trying to replace X without implementing a lot of shit X did. You could debate whether X should have been doing those things instead of other components, but the end-result is Wayland as a protocol does too little to replace X and therefore can't.
It's a fundamental difference in the way its run. X11 is mechanism over policy while wayland is policy over mechanism. I remember when VR was hot stuff and x11 got full VR support in 2 months, for wayland it took 2 years and gnome wayland still doesn't support it.
The original dev left the project looong time ago. The people who continued it mainly used it for car dashboards so its really very barebones to ger stuff to screen. Then gnome picked it up, but instead of creating wayland protocols for desktop use, they simple use dbus or other gnome vendor-locked way to do desktop things. Wayland becoming barely usable is the works of volunteers from wlc and wlroots, coming up with some actual extensions to try and standardize shit. The core wayland project however still is utter shitshow, by having strange defaults in the WSI implementation that fricks up games and apps codes in the traditional way, and needing hacks for to avoid crashing on stalls.
>all these hateboners about wayland vs xorg from idiots who have never coded shit
Kek.
How do I even begin to gather the knowledge to contribute to these projects. Everything just sounds so fricking cool.
X11 - Gayland0
Xorg is dying every day.
Wayland is incomplete and pretty much is in the state Xfree86 had in 1999.
Meanwhile, arcan autists still don't produce shit.
The entire display server stack in Linux is broken. It's over.
Works on my machine (xorg that is).
Digits of truth
checked
hope is the last thing one ever lose when it comes to linux
still using windows 7
>valve dev
>skill issue
Name an iconic duo
Gayland is bad but this dev is even worse
Seriously how did no one critique this troony holo live dev profile picture
>its okay if the troony supports my X11 KDE over the others wayland gnome
>dubs
>checked
Dubs of truth
why dont we just use surfaceflinger
That's what Mir originally was, basically a PC port of surfaceflinger. Red Hat unleashed the FUDgates and threw a bunch of money towards Wayland in horror because there might have been a part of the Linux ecosystem that they didn't control.
>the 30 thumb downs from assblasted gayland homosexuals and devs
how are they the future, again?
Want good wayland gaming performance, just use gamescope. It's got a nice steam integration, and if you're using hyprland or sway you should be using it anyways. It is a better compositor for games.
Also, keep an eye on KDE. Last months they already have stolen the spotlight of GNOME in terms of Wayland development.
Meanwhile, GNOME devs are crying because they were called slow :^(
Have you seen the state of KDE 6 + Wayland? It's a fricking disaster.
this is literally the Gamescope dev himself telling SDL they are not ready tho
>30 Black folk give the thumbs down even though the dev who had added israeliteland support explains why it won't work
I hate updoot trannies and will everyone with my bare hands if given the change
a political system that give everyone a voice is garbage because 99.99999% don't deserve to voice their worthless consoomer cattle opinion
The problem with updoot/downvoots is that they provide a means of commenting on a situation without any critical thought whatsoever.
At least to make an idiotic post you need to go through the motions of trying to put it into words, which provides the moron at least an opportunity to recognize their own moronation. Some morons can't even muster up the effort, and will simply remain silent.
What's the idea of wayland and what it tries to solve that X doesn't?
supposedly it's unmaintainable code without proper vsync support and horrible latency or something like that
Joshua-Ashton, Stenzek, WM4
they all have warned us
>Stenzek
as in the guy from Duckstation and PCSX2?
homie is this really worth your time? Arguing about shit nobody uses, and nobody will remember in 50 years? Over tiny irrelevant shit. Get a fricking job dude
don't voice your worthless opinion, wagecuck
now get back to work
You have no idea how little my time is worth back the frick off
I'd really like to upvote this comment but this website doesn't seem to have a vote feature. How odd!
Are games not gonna run on wayland?
They don't anyway. The run on xwayland through wine.
Oh, so I won't get bug reports from wayland users b***hing if I'm using sdl?
i love sdl so much bros, can't wait for sdl3.2
of all the competition, slouken and icculus are the ones i trust to make a solid graphics API wrapper
Gayland is just sabotage, it will never be usable
At the end of the day, is xorg that messy to create the need for this?
What is it like 400,000 lines of code? 200,000 for the Linux specific stuff? I've look at the code (not recently), and it's pretty bad.
I was looking at the xserver code the past few days.
It's actually fine. It's your typical generic 80s/90s C code meshed together with the meson built system.
Very simple actually, not hard to understand at all.
>Very simple actually, not hard to understand at all.
Oh well shit go fix xorg for us then smart guy, what are you waiting for?
What needs fixing really?
Maybe HDR support. But this needs a protocol extension and requires working together with multiple parties like e.g. Wine to make sense.
I dont think it needs a protocol extension. The way hdr is done on windows these days is without an api. You create the window with 10/12bit depth and that signals that it's hdr aware. If you set the linux drm hdr metadata in the x11 modesetting driver then such applications will automatically be displayed with hdr when they have pixels outside the 0 to 1 color range. The only thing you really need to support for is tonemapping for non hdr applications and that would be the job of the compositor. The compositor already has the information of color format for windows with other metadata as well so nothing really needs to change protocol wise I believe.
> that signals that it's hdr aware.
This "signal" needs to be a standardized protocol extension.
It's really not much to bring HDR to X11. It just needs to be coordinated and standardized in some way.
The main customer will be Wine. So they have to be on board as well.
I forgot to mention that the hdr metadata needs to be exposed to the compositor, but that can be done with per monitor properties which is already done for edid for example. And also if you really need to add extra data to implement this it can be done with window properties and monitor properties, no need to modify the protocol.
BTW, chromium on x11 is already aware that it can display with hdr. If you enable 10 bit colors for x11 and open youtube in chromium it will say that it's displaying with hdr so just enabling hdr in xorg should make chromium hdr work. This is only on x11, not wayland.
why is UI such a big problem for linux?
Sabotage by Microsoft.
The next sabotage project in full swing.
I'm gonna start learning rust since you morons seem to be wrong about everything
and that's a good thing!
>pulseaudio and systemd were made by microsoft plant
>wayland and rust are endorsed by microsoft and they clearly pay for the shilling
grim
There's rust in the windows kernel too now
nah rust is legitimately a cancer of a language
to be fair I don't care if the programmers writing it suffer if it prevents le memory leeks
The power of open sores, they still can't agree on a standard graphics library. LOL
Everybody already agreed on X11.
Then came Wayland and fricked everything up.
And Wayland is 100% corporate sponsored and "open source" in name only.
How quickly would a superior alternative take to get to the point Wayland is now? Or would it be better to just fork it, circumventing whichever roadblocks it has and overhauling poor design?
> How quickly would a superior alternative take to get to the point Wayland is now?
It already exists. It's called X11.
X11 alternative.
Wayland would have been ready now if not for years of refusing to add anything at all to it.
And then requiring months long discussions to add even the most obviously needed features.
Why hasn't anyone meaningfully forked it? Politicizable stigma?
> Why hasn't anyone meaningfully forked it?
Because X11 just works and everybody simply switches back to it quickly after trying Wayland.
Probably cause wayland is not that appealing if you expect to have full fledged DEs using it.
Wayland + wlroots + a billion portals might be, but at that point it might be better to learn from x11 and wayland errors and make a better alternative.
Listen I am not gonna update my Linux distro until this mess is cleaned up. I will stay on xorg forever if that is necessary.
So sort your shit out FFS.
Gayland btfo
As it turns out, every single bizarre, hipster design decision that wayland made was fricking stupid and needs to be hacked around.
This is just in:
> https://github.com/libsdl-org/SDL/pull/9345#issuecomment-2023112550
> WONTFIX!
> There is no use case!
> Wayland stays the default!
Genuine mental illness.
>it doesn't work
>but it might work
>the 15 year old vision of the future is more important than the present
>It sounds like the Wayland folks are aware of the issues and are making progress towards addressing them. I'll leave Wayland default for now while they are doing that, so it's easier for people to test and provide them feedback. We will re-evaluate this as we get closer to full release.
I don't know what you're getting at here. Everything I'm reading in that thread tells me it's being handled.
>it's being handled.
The issues the valve dev is talking about has been a problem ever since wayland was conceived and people have talked about it for multiple years. But sure two more weeks.
You literally linked to a comment where they said it was being handled and they're going to keep Wayland as the default.
>You literally linked
wasn't me bud
Okay, then why did you reply? To wax philosophical about bike-shedding or to discuss the technical implementation? Clearly it wasn't the latter.
I replied to tell you that it's not being handled in any meaningful way. It's not complicated.
I'll trust the the word of the SDL developer and not some random anon on a sheep-herding forum. The protocol has improved a lot over the past 3 years, the bike-shedding problem isn't anywhere close to prevalent now.
The core protocol has not meaningfully improved at all over the past 3 years. What you meant to say was that they added more hacks around its misdesigns.
>the core protocol
Which part do you think hasn't improved specifically? Are they working towards a resolution? If they aren't, show how.
No, he's right. The core protocol has barely changed in 3 years or 10 years. There are simply more testing protocols that compositors can implement, to get the same features they have on xorg.
Developers say otherwise.
>They haven't exposed positioning to clients.
Giving clients complete freedom to choose geometry and index order leads to moronic shit happening like destroying a dock. It's not necessary for any kind of idiom. If you want windows positioned exactly, you tell your compositor to do it.
>Developers say otherwise
I don't know what kind of moronic cope this is but anyone can go see how few wayland protocols are stable, i.e. mandatory to implement and uniformly implemented.
https://wayland.app/protocols/
Many of them are capable enough of being used and won't change significantly.
That's not how that works. You need a protocol if the client has to make a request, the compositor/server can do what it wants.
>capable enough
>won't change significantly
lmao. Let's just scrap the stable tag altogether then. Clearly that's becoming a reputational issue for wayland too.
And the client has to make a request.
The compositor doesn't know how it should be positioned when for example I want my application to remember the position and size of an auxiliary window and restore it to the same size the user left it before.
> I want my application to remember the position and size of an auxiliary window and restore it to the same size the user left it before.
Auxiliary windows should be handled through layer surfaces which you can position absolutely although in practice this has become very rare (most programs that do require absolutes are rendering a canvas). You should not have control over where top layer windows are positioned. There are no idioms that require this. If a user wants to have a window positioned in a particular way the compositor can expose the ability for the user to control that.
The compositor has no idea what the window is and no idea how it should be positioned.
This is a really basic idiom that every desktop had since the beginning of time.
Even Amiga Workbench from the 80s would remember the last position of every window for every different open folder.
The compositor shouldn't, the user does. If the compositor wants to expose the ability for the user to control exact window positioning it can do that.
>This is a really basic idiom
It's not an idiom.
The compositor can't do that.
It doesn't know exactly what the window does and represents.
Only the application does.
Are you saying that users don't know what their windows are?
Do you want users to manually position their windows to where they like it every single time again?
I absolutely loathe applications that don't remember window positions.
If this is a desired feature you could expose it through a GUI with rules based on instance/class names or through scripting, dbus, etc.
let's overcomplicate everything with an extra dbus service and do it all in a round-about way because Wayland refuses to support an extremely basic feature every fricking desktop in the world had since the beginning of time.
I can't believe some people don't like Wayland. Such a great piece of software in such great hands.
Compositors already expose a dbus interface.
>basic feature
Unnecessary security regression.
>have some insane hack instead of supporting the feature
every time. what's wrong with the minds of waytards?
>hack
Exposing the ability for users to position windows is very literally a core part of any compositor's job (which targets a user-facing environment and not, ex. a kiosk). Doing so is definitely not a "hack".
The sane way to implement that is using the window API not some unrelated IPC
>The sane way to implement that is using the window API
A desktop environment can do this through GUIs, and then use its own API. Smaller compositors do not provide GUIs, you script them, and if you want to provide users the ability to do that easily you use dbus or some other form of IPC along with a tool that can handle dispatch conveniently. Some of them do expose a proper API and target programmers. These are all perfectly reasonable solutions and they are all very sane.
This is completely irrelevant babbling. The correct API to use when implementing a feature like this is the same one the display server actually uses to manage its windows. Everything else is just a hack.
You seem to fundamentally misunderstand that the compositor is already doing that. They're just exposing an easier interface for the user. This is exactly what many X window managers do.
The correct interface is the window API.
So you're proposing that every user write their own compositor or window manager for X?
Every user should be able to communicate what they want to the compositor using wayland APIs. It's not complicated.
That's not how it was handled on X in the majority of cases. Most tiling or even small stacking window managers exposed their own API or communicated through another interface. Clients being able to position themselves isn't required for any kind of idiom, nor is it required for users to position windows, compositors are what do that.
I don't care about how you think it worked on X.
>Clients being able to position themselves isn't required for any kind of idiom
Many applications work like this and they work fine everywhere except wayland because it's a crippled platform.
I don't care, it's a security regression.
How do you handle bounds checking?
How do you make sure the client doesn't wipe out a dock?
How do you handle any other form of collision?
In order to do this the client needs a lot of information, information it shouldn't have.
Yeah, rejecting an invalid position is absolutely impossible. No one can do that except every other display server.
>hey server, I want to be here
>okay, there's something there
>okay, but how do I know where I'm supposed to go now?
>oh, sure, here, let me give you detailed information about what's on screen right now so you can position yourself properly
I know this is impossible for you to conceive but the compositor can just ignore the request if it's programmed by some paranoid schizo.
So what does the client do? How do you then handle a standard for how the window should be placed afterwards? What if there's already something there? What if the fallback of the fallback also has something. It's just a bunch of moronic cases that need to be handled because some schizo thinks his client should be a window manager
>So what does the client do?
Nothing because by your own admission it shouldn't know any of this information in this hypothetical.
>nothing
So now the compositor handles it anyways, and a feature that should work in the client doesn't actually work, which is not only bad for the user but creates an unnecessarily stupid implementation where the developer doesn't even know if his request succeeded.
>compositor handles it anyways
No it does nothing in this example other than whatever the default for window opening is.
>whatever the default for window opening is.
So we're in agreement, it's being handled by the compositor after that. Imagine using a protocol or an API that doesn't return anything after a request. What a fricking moronic conception.
>Imagine using a protocol or an API that doesn't return anything after a request. What a fricking moronic conception.
This already described wayland before they added window capabilities. The thing we are discussing here could just be added to that. You blatantly don't understand the any of this.
That is not how the protocol works.
Wrong, that IS how it works. If the compositor doesn't implement wm_capabilities, there is no way to know if minimize works or not.
>if the compositor doesn't implement the protocol it doesn't work
No shit?
Wrong again. minimizing is a part of xdg shell version 1 which all desktop compositors support and use. wm_capabilities is a subset of that which was added later. Isn't it funny how wayland advocates know nothing about the thing they love so much?
Wayland had to implement xdg-shell-capabilities you moron. If you don't support that part of the protocol it doesn't work.
also*
Who is "wayland"? Did you mean compositors? Let me refresh your memory here. You said:
>Imagine using a protocol or an API that doesn't return anything after a request. What a fricking moronic conception.
To reiterate, this described wayland for many years. For example if your client sent a minimize request, there was no way to know if it worked because the API doesn't return anything at any point. They later realized this was a bit moronic and added wm_capabilities which allows the compositor to advertise support for various xdg shell requests to the client. It is completely possible for a compositor to only support an older version of xdg shell so the "doesn't return anything after a request" part could still be true today.
Yes and it was 100% moronic at the time. I don't care about what Wayland was like 4+ years ago. I care about what it is now, what's being worked on now, and how it's progressing.
You demonstrated multiple times in this thread that you understand nothing about the details of wayland. Why have such a strong opinion on it?
No, the only thing I demonstrated was that you're a moron who deliberately misunderstands what people tell you to try and build an argument off of it once you get backed into a corner.
No one is buying that. The reason why the suggestion of a compositor hypothetically doing nothing for window positioning triggered you was because you didn't know that compositors can already advertise support via wm_capabilities. That's also you made that very funny statement without realizing it also applied to wayland. I presume you spent the past hour or so googling it and doing your research to avoid further embarrassment. Until next thread.
You made up a bunch of total guff, said I assumed many things I didn't even discuss, then came up with some fantasy land lie about my argument. You're so buttflustered right now you can't even write properly either.
>You made up a bunch of total guff
I'll even link you the MR. It's all XML. Even a no-coder like yourself can understand it.
https://gitlab.freedesktop.org/wayland/wayland-protocols/-/merge_requests/122
That's not even what I'm talking about. You basically injected an entire argument into my posts that I didn't make.
This is coming from the guy who typed "Wayland had to implement xdg-shell-capabilities" as if that made any sense lmao
Yes, moron, xdg-shell is a part of the fricking protocol.
>xdg-shell is a part of the fricking protocol.
Wrong again. Why can't wayland advocates just take two seconds to google it lmao. xdg shell is a separate protocol on its own. It's not "a part of the fricking protocol" whatever you meant by that. Certainly not a part of the core protocol.
>xdg shell is a separate protocol on its own
I'm not saying that part of the protocol IS xdg-shell you complete autist
This just reeks of ESL. What part of the protocol? Which protocol? Are you attempting to say that wm_capabilities IS xdg-shell? That would be incorrect. It is only a small subset of it. Please, express yourself in an non-ambiguous manner so I don't have to guess too much when mocking you.
>Are you attempting to say that wm_capabilities IS xdg-shell?
There you go again, shoving words into other people's mouths, making arguments out of nothing. Holy frick.
Instead of using a myriad of ambiguous pronouns and getting upset that people misinterpret you, why don't you actually type what you're trying to say clearly? I put a question mark for a reason because I wasn't sure what you were trying to say there and now I really have no idea what you meant. I'm sure it's something stupid though.
wm_capabilities is just an event that tells clients what won't or will work in the context of xdg-shell support (for the compositor) and only provides very basic information about what a top level can do. It's not the same as firing and then getting _nothing_ in return which is what would be required to maintain security in the context of a client positioning itself. The protocol shouldn't have a baked in security regression.
Literally just add a MY_TOPLEVEL_WONT_BE_POSITIONED enum in addition to the API that makes a request for a position. That's all you have to do.
No compositor should be allowing clients to position themselves. There's literally zero purpose to having that as an event. Adding it doesn't solve the core problem.
Oh after all this inane arguing we just looped back to the classic "you don't need that feature". lmao every fricking time with you chucklefricks
Your feature is a security regression. Go chop off your dick and try to turn it into a vegana, I'm sure that's an appealing "feature" that everyone should be allowed to use.
>muh security theater
way to prove his point
If you don't care about security keep using X.
X has been designed and used in actual multi-server users, it has much more proven security than Wayland.
Just because it doesn't try to protect a user from himself like an iPhone doesn't make it insecure.
>used in actual multi-server users
You mean multi-user servers?
Yeah, got distracted and swapped the words.
Someone clearly forgot their anti-psychotics this morning.
Why is waytrannies only argument that "LE NOONE WHO CRITICIZES WAYLAND UNDERSTANDS IT"
It reminds me when I complain to the moron at work who maintains the program everybody uses. "NO I CAN'T DO THAT DON'T YOU UNDERSTAND, JUST DO THIS"
He is inevitably fired 6 months later and it repeats every 3-4 years. What is the name for this?
Literally no one said that.
Basically nobody is going to take the time to learn your shitware because it is heading towards a full rewrite. So if it seems like "nobody critiquing wayland understands wayland" then that is why.
No one said that either.
Are you in college, man?
Are you schizophrenic?
The act of getting a rejection lets the client know about something it shouldn't. It's a totally unnecessary feature for clients to have that compositors should handle.
>muh placing windows on the screen is a security regression!!!!!
literal mental illness
>dbus
So it really was a redhat ploy from the get go. Absolutely shameless.
>redhat
Redhat devs have adhd teenage developer syndrome. They couldn't maintain something as simple as a display server so they had to remake the wheel while leaving the one everybody uses to rot. They spent 16 years remaking the wheel, leaving out tons of basic functionality for no good reason.
The best reason they can give for ditching xorg is that it is old and unmaintained.
YOU WERE THE ONES WHO DECIDED HOW MUCH MAINTENANCE XORG GOT!!! WHAT TO SAY YOU WON'T OR ALREADY HAVEN'T GIVEN UP ON WAYLAND!
>If you want windows positioned exactly, you tell your compositor to do it.
Which requires a protocol that governs how you tell your compositor how to do it.
They didn't deprecate integer scaling like they should have done. They haven't exposed positioning to clients. They haven't fixed frame callbacks being utterly moronic. They haven't added any kind of access controls. etc. and etc.
> it's being handled.
It's being "handled" for 16 years.
That's what people are "getting at".
Well at least they're supposedly going to check for protocol existence. Note that both are merely PRs and are in dev hell but oh well I guess.
you gotta be kiddin me
why is everyone getting on their knees to suck on gayland's wiener?
Linuxtards are really something else. It's like they don't want people to use their shitty OS.
Microsoft doesn't want people to use this OS.
That's why they are investing in sabotage.
but we're resisting it
right, IQfytards?...
we are too busy creating advanced algorithms to parse the windows registry efficiently
>hmm these wayland trannies seem trustworthy
>I bet they are actually working on it and not prolapsing their buttholes with giant dildos all day
grim
Is steam os still a thing?
Somone made a liveCD installer for it called Holo OS. I've tried it out and I thought it worked great. Ended up being the only way I could even play a recent game, it would stutter and freeze on Windows. Catch is you need an AMD GPU, and obviously it's the same immutable steampozzed arch.
That homosexual with the conan profile picture really needs to go touch some grass holy frick, and calling users "stakeholders" is so disgusting. It's a tool that needs refinement, not a movement that should be the defacto. x11 works. Sorry wayland troons.
>making a windowing system
>completely shits the bed if the client has multiple windows
this is the state of wayland 16 years in btw
gayland tards are actually going to argue that you literally can't do things that people already do on xorg o i am laffin
i'd have more respect for them if they just admitted they don't care and are getting paid to develop a thing they don't even think is good
>X-tard has argued himself so far into a fricking hole that he's now saying that the only acceptable solution for window management is everyone writing their own window manager
Looney land.
NTA but that's only slightly more insane than the Wayland status quo, at least making an X11 WM is easy.
wayland is worse at being a desktop environment than the homosexuals who make the shit are at actually being a women
He already folded guys
This is a perfectly fair compromise, SDL users get an actually functioning library and gayland tards get their illusion of progress. Kudos to the valve wagie.
What's your gripe here? This was the obvious, and more importantly, correct, resolution to this issue. SDL, X11, Wayland, end users, programmers, everyone came out of this where they wanted to be.
The correct resolution to this issue was getting rid of the fundamentally broken trash that is wayland
Thanks for the response, I just wanted to clarify that you're mentally ill. Wishing you the best, though.
Apologist for utter garbage detected
Don't (You) me
Wayland mental illness won in the end
>if the protocol that we need is available we will use native Wayland, otherwise we won't
>this is checked at runtime
Sounds sensible to me
LOL
ITT: We predict if/when these protocols get merged.
Betting on 4 years.
I believe less and less in free software. It was one of thing when I was a kid and Microsoft was EEEVVVIILLLL and Linux was exciting like the wild west, but now I just see people fighting for minutiae all the time. It's like democracy.
Arcan solves that problem.
Literally what
A shit Xorg clone made by a deranged individual that isn't has funny as Terry Davis.
So this is the angle you're going with? Just call the guy a schizo because he uses neovim?
No, he isn't a deranged individual because he uses a text editor but he is because his writings are stream of consciousness with added technical jargon.
>he's deranged because I don't like his blogposts
yeah okay bud. I'm sure you're the paragon of excellent mental health yourself.
Arcan ain't gonna happen.
It's either endless idelogical debate or you just get what microsoft and apple give you.
Applications are not window managers, they aren't compositors. Those are tools built to handle that task, they are the ones that should be handling it. Your application does not need and should not have window management capability.
>valve and Redhat forcing technology
>Unemployed incels raging
Simply go somewhere else unemployed plumber.
I enjoy my Wayland Nvidia Gnome Steam Desktop with no extensions to its fullest.
Show me a screenshot of cyberpunk with hdr enabled and I'll be impressed.
Wait until Cosmic comes out deprecated homosexual.
the absolute state
I use i3 and Xorg, why should i care?
>desktop linux finally gaining marketshare
>wayland suddenly starts getting pushed hard and wreaking havoc
really makes you think
Honestly if freedesktop and redhat never existed linux would be in a much better state. Nowadays it's just windows-lite.
RedHat does seem to have brought nothing but evil to Linux. Valve is basically what RedHat pretends to be.
>"toplevel", for your top-level application windows, and "popup", for things like context menus, dropdown menus, tooltips, and so on - which are children of top-level windows
OOP is the reason display servers will never be cheap, correct or performant.
Idk what any of this means, anon. Could you write me an 18 page essay explaining it in extensive detail?
>gayland's so bad that anyone that argues about it sounds like they have no clue what they're talking about
every time
>casually take a gander at wayland repo
>expect to see some graphics related stuff
>it's all multithreaded networking code
Neat.
>there are people seriously advocating for Arcan
The guy who wrote is so deranged that most of people don't understand how exactly Arcan works. Instead of documentation they get mumbo jumbo and rants.
Those who seemed to have cracked it says is basically Xorg again.
He's perfectly willing to answer questions if you ask him.
Wow!
Great documentation!
10/10
I mean say he goes out and documents a bunch of shit but it's still in alpha and then he changes it maybe in some dramatic way
What's the point of wasting all that effort
I agree. Why waste effort in Arcan?
We could could be using that manpower into fixing some glaring Xorg design issues, like it's incapacity to do proper multi monitor with different refresh rate and DPI.
Arcan is worked on by literally one guy. Nobody's manpower is being wasted. I guess in theory he could hostile fork xorg to "fix it" maybe, but you really can't blame him for not wanting to go through all that and just rolling his own solution.
His solution is significantly better in many ways.
Is Xorg 2.0 with extra autismo and worse performance network capability.
Networking capability in Arcan is, if he follows through, going to also be significantly faster and better than using ssh.
Also the TUI shit is some terry level improvement on CLI, guy is a certified genius
Reinventing x.org, but worse this time, is moronic not genius.
>sqlite3, openal-soft, sdl2, opengl, luajit, gbm, kms, freetype, harfbuzz libxkbcommon
LOL
Damn near everything about is (planned) to be better.
Funny enough, he has a fork. Showed re-using X11 WMs in the last release post. Seem to be working on screen sharing compatibility:
https://github.com/letoram/xarcan/commit/c59643bceebc20913f61a697adcff87a00784fbd
Wayland doc on security:
https://wayland.freedesktop.org/docs/html/ch04.html#sect-Protocol-Security-and-Authentication
Arcan on security:
https://github.com/letoram/arcan/wiki/Engine-Security
https://arcan-fe.com/2023/06/15/the-quest-for-a-secure-and-accessible-desktop/
Who looks like they are actually trying to document what they are doing?
Why don't we just use http(s) or something for all file communication? No weird special cases, games just work, in or out of the browser
You mean tcp? Why not just use 9p for everything like god intended?
because it's a bloated fricking protocol
also can't do multicasting
because it wastes tons of resources. you don't notice it when its used for throwing around little json files but its terrible for heavy use.
>that pfp
YWNBAW
Someone already fixed that and created a patch for it.
Red Hat refuses to merge it.
I'll stick with 9p/ip running my custom display server thank you very much
Why do fricktards keep bringing up arcan? Nobody asked.
It's a thread about X vs Wayland, now that Red Hat has officially given X an EoS date, it's time to start thinking about alternatives, because trust me, it's going nowhere without RH's devs.
Any modern project that uses a scripting langauge like lua is doa. Nodevs don't understand. Take it away jblow
Didn't that guy blow like 20 mil on his language and it's still not done?
no he jblow
Straw man. He made 6 points in that video on why scripting languages are a terrible idea.
ARCAN IS DOA. We both know it. Stop trolling.
Has blow developed a game as big as WoW, what about an application as widely used as neovim? What about even something like Project Zomboid, which has a massive community of modders thanks to lua. It's not like Arcan is entirely in Lua.
And webbrowsers use js.
I'm glad that at least some people know better. Don't justify a shitty situation by saying "it's shitty here too."
The point of having lua in there is for people who don't have the experience or are just hacking around. You don't need to or have to target Arcan with Lua (and most people don't)
It's not just lua, there is also sql there for some reason. This is giving hard "video game made in 2003" energy. Is this stepmania?
runtime config is consolidated into a proper database
Can you define "proper database" I've never heard that in my life. Seems like you are just religiously defending bloated nonsense.
Are you a troony still mad about Saint Benedict's rule?
>bloated nonsense
Storing all of your configuration information into a database is way less bloated than having it scattered across your system in text files. It also makes migration insanely easy without having to resort to systemd-home tiers of bullshit.
That is literally counterfactual. A filesystem is by definite less bloat than a filesystem + unnecessary opaque file-system.
Sneed
>A filesystem is by definite than a
No it's not. Also you probably meant "definition" ESL. Having a key-value store in some kind of database is literally something we've been doing longer than you've been alive precisely because filesystems are so shit at handling it.
It's literally a relic that won't die. I love when cruftards think they are the future
>no the dev can't just write a program, I sneed a database that the user queries so they can write a shitty lua script that
LMAO
>write a lua script
Where does lua come into all of this?
>a relic that won't die
Only a neet nocoder would think so.
Again. This thread is not about arcan. Noone uses arcan, nobody will ever use arcan. Anybody and their mother could tie together a dozen disparate technologies to make morons like you clap their hands. Anybody with more than 1 year of experience knows to run when they see an open source project with lua/sqlite/cmake the holy trinity of shitware.
>lua/sqlite/cmake
>not js/tauri/rustc
the frick is wrong with sqlite?
Nothing he's a pseud 1st-year who thinks john blow is god
sqlite3 is probably the safest 'FILE' (as in fopen("mycfg", "w+") abstraction that doesn't shit itself on concurrent access or partial write because crash/reboot/power outage.
that guy is a moron, those three things are some of the best software ever created.
>arcan_db show_appl durden
>arcan_db add_appl_kv durden shadow_on true
Do you see lua here or do you see a c program running inside of a shell. Let me know when you're ready to stop being moronic.
Now you might be like, "why? just surf through 8000 folders with 80 different configuration formats", I think "oh my god I can query configs across my entire system and it will only take .1s"
Oh, you mean "I don't know how to use unix cause I spent my 20s learning sql"
If you think using find to go through literally hundreds of thousands of files is going to be faster than just querying a sqlite db you are genuinely a moronic nocoder
anon...
At any rate if you have that kind of configuration problem with your program the developer should write an actual program, a configuration tool, that solves it.
>with your program
It's not just one program, you moron. At least read about the server before you shit your idiotic opinion all over the place.
>no I can't just symlink relavant config files to a common directory. I SNEED A DATABASE
symlinking doesn't solve the moronic issue of configuration files being all over the damn place, like oh boy now not only do I have some buttholes config directory outside of .config I have a symlink to it in .config too! WOW
>lua on a display server
god has abandoned us
Blow would get a heart attack if he ever were to look at freedesktop.org crap
luajit really isn't that slow though
Your brain is fragmented.
If RH doesn't want to support Arcan, arcan would keep being a passion project with at best, support for some embedded niche use case.
We could have got NeWS or Display PostScript but UNIX autismo wanted to use that VAX shit.
>NeWS or Display PostScript
completely obsolete with modern graphics accelerated displays
They were still a saner option than X11 and it wouldn't had take a while to make them speak modern GPU stack.
Because it's not ancient crap with over a decade of legacy cruft and hacks like Wayland.
The Linux desktop is going to keep being a joke unless it finally gets a modern display server.
I love watching Linux streamers. You can see their shit tear non-stop whenever they traverse text files in editors lol.
No idea.
Imagine wanting to run Xorg 2.0 in order to display wayland clients anyways.
>why do people talk about a display server in a thread about display servers
truly a mystery
All this talk about millisecond shit. How do you even measure this? Vulkan doesn't provide you an exact timestamp of the framebuffer swap, right?
Or do graphics programmers shoot their monitors with high-speed cameras to debug shit?
Are you talking about presentation timestamps? You're correct that vulkan doesn't provide this. That information comes from the kernel/drivers. How you know it's actually accurate? Well I dunno; just trust that they didn't frick up I guess.
All right, thanks.
In fact, to my knowledge, a lot of what has been created for Arcan is in Zig, or C, or C++.
Call me old fashion but I expect a display server to build it I type make in any system.
Arcan has a bunch of C APIs. It's not just lua.
I just watched the latest arcan video.
The dev appears to be a random loner living in his moms apartment.
He literally uses vis, the meme sam-inspired text editor that liters directories with hidden files. Need I say more? Arcan is a nothingburger.
that's... neovim.. not exactly a controversial choice. Dev has his on UI driver for it ..
https://github.com/letoram/nvim-arcan
>local man writes code in neovim how dare he
lmao really
He's a security analyst, and a well regarded one.
Ah, a jerking off monkey.
so, the average IQfy poster