Global shortcuts on Wayland apparently require using fucking D-Bus and the xdg-desktop-portal disaster. How did anyone think that was a good idea? It's so bad that 3 years later, adoption is basically none.Sure, wayland protocols suck, but this is so much worse in every possible way.It's no wonder that the native GUI situation on Linux is a disaster. The simple and effective APIs are "deprecated" and replaced with utter shit like this every 5 years, because keeping existing programs working is too much of a "burden" on Redhat.The protocol is here if you want to see how overcomplicated it is: https://flatpak.github.io/xdg-desktop-portal/docs/doc-org.freedesktop.portal.GlobalShortcuts.htmlThis is literally a single function call on Win32 and X11, btw.
>>109143222what is this global shortcut? Like hotkeys a window can listen without being in focus? why they didn't have a why to allow some programs this privilege? It feels like wayland designers didn't have an adult in the room to tell them their ideas are shit, probably some tranny echo chamber where you're not allowed to disagree with the cabal.
>>109143483>Like hotkeys a window can listen without being in focus?Yes, exactly this. Commonly used by screen recorders, password managers, etc.https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-registerhotkey>why they didn't have a way to allow some programs this privilege?Your guess is as good as mine.
>>109143222Baroque and bloated design is a sign of failure and lack of clear "vision". There's no other way to put this.
So I can talk about this a little bit now because of how immersed I have been in GNU/Linux development for the last while.Dbus is fine. You use Dbus, yes you do. You're a fucking liar if you say you don't use dbus. Either that or you hate yourself and you hate me as a developer. There is nothing wrong with it, it exists to make everyone's life easier. It takes minimal resources.Now if you wanna bitch about things like bonjour services and other GNOME daemons that tend to float around, yeah those suck.The native GUI situation on GNU/Linux is a dichotomy between GTK and Qt framework. I've always been on the GTK side. Personal preference.>Simple and effective being deprecatedNot anytime soon, GTK3 is going to be around basically for another 15 years. GTK4's biggest improvement is in GPU acceleration and async, both of which can be done (although it is harder) in GTK3.The XDG desktop portal specification however is one of the best things I've ever seen. I'm currently working on a screenshot tool (a fork of shutter, modernized) and plan on implementing Wayland support, as well as scrolling screenshots, as these have been a heavily requested feature. My plan is to use the portals API along with heavy flatpak sandboxing, to assure user safety and privacy.
>>109143593cooming to lolies rotted your brain, anon.
>>109143644Please stop responding to me if you don't have anything constructive to say. Thanks. I don't masturbate to lolicon because I'm not a pedophile. Kindly fuck off.
Hyprland supports global keybinds and I just made my own DE in quickshell. I do not care anymore.
>>109143593How is hundreds of lines of unnecessary boilerplate "one of the best things you've ever seen"? This garbage is why the year of the Linux desktop is never coming.https://github.com/ghostty-org/ghostty/blob/main/src/apprt/gtk/class/global_shortcuts.zigOne function call in X11, btw.>Dbus is fine. You use Dbus, yes you do. You're a fucking liar if you say you don't use dbus. Either that or you hate yourself and you hate me as a developer. There is nothing wrong with it, it exists to make everyone's life easier. It takes minimal resources.There's no good client library for it that isn't linked to GNOME, systemd, or is just plain garbage (libdbus).>The native GUI situation on GNU/Linux is a dichotomy between GTK and Qt framework.And that's a problem. There should be a platform API that never breaks compatibility, and both of those do it regularly. As long as Linux keeps breaking old programs, it'll never win.>My plan is to use the portals API along with heavy flatpak sandboxing, to assure user (((safety))) and privacy.Meanwhile on Windows, they download the .exe and it'll just werk for the next 20 years.
>>109143716>Muh boilerplateI don't care about that, I program with agents>LInux desktop is never comingI don't care about winning a popularity contest>One function call in X11Simpler doesn't always mean better. I prefer Wayland for its security posture.>No good client for it...It's built on top of sockets smfh...>You can never break compat!!!Yes we can. It's a good thing that we don't have to be chained to perfect backwards compatibility, that is a Windows thing. Leave your Windows pattern elsewhere. We don't want to support 30 years of bullshit in GNOME, KDE, XFCE, etc. if you want that bridge gapped then pack an appimage, it was made for that.>As long as Linux keeps breaking old programs, it'll never win.It doesn't break. You can just not upgrade or do what I said above.>Meanwhile on Windows, they download the .exe and it'll just werk for the next 20 years.Good for them. I don't care about Windows.
>>109143651nta but I think your original reply didn't have anything constructive to say either, so I think >>109143644 made a good and accurate response
>>109143593>>109143768>vibecoded wayland gtk flatpak-only screenshot toolI will not be using your software.
>>109143768>I program with agents>I prefer Wayland for its security posture.>It's a good thing that we don't have to be chained to perfect backwards compatibility, that is a Windows thing.Kill yourself.Saboteurs like you are why the year of the linux desktop is never coming.
>>109143768>I prefer Wayland for its security posture>what is xnamespacelet me guess, you use rustdon't ask how do I know, it's magick
>>109143593Love dbus, gives me the tools I need to find programs using dbus so i can disable them.
>>109144402Based
ziggers whine about everything
>>109143593dbus is trashit literally never worked for me when i used any wmthere were always some fucking errors in logs WHILE there were no visible problems on the application sideif applications work without this shit why the fuck would i want some bullshit thing that fucking uses xml payloads over network socket for IPC instead of using fucking shared memory with binary data?
>>109143483>why they didn't have a why to allow some programs this privilege? because you can either have features or "security", and youre on the wrong side of history chud
>oh noooo i have to use an asynchronous d-bus handshake to bind my precious mute button in discord instead of letting every single shady python script from the aur read my banking passwords via xgrabkey, linux is literally ruined redhat killed my dog
>>109143716>And that's a problem. There should be a platform API that never breaks compatibility, and both of those do it regularly. As long as Linux keeps breaking old programs, it'll never win.In w32 terms those are GDI+ and Direct2D + Directwrite, as in level that lets you draw shapes color and text and on top of which GUI libraries/ frameworks then build. And yes, there is no equivalent in linux, which is 50% reason it sucks in desktop use. https://learn.microsoft.com/en-us/windows/win32/learnwin32/overview-of-the-windows-graphics-architectureOther 40% is next level down from it that being set buffer pipeline that currently in Windows lets it no problem draw mixed SDR+HDR content no matter what screen refresh rate / resolution. And where last point stuff then dump whatever it's rendering. https://learn.microsoft.com/en-us/windows/win32/direct3darticles/high-dynamic-range.And last 10% is same for audio.
what do you expect from the bird language like this:> Portal for managing global shortcutsobviously author was a no-coder midwit at least and a retard with a high probability.
>>109146466Bot post.
>>109145594>xml payloads over network socket for IPC instead ofbecause its GNUnix way of binary retardation, they need <stdio.h> included in every sourcefile. everything must be split into chaotic binary blobs costing $50,000 dorrars each, hehehe.no holistic approach, in short. thats why everything has some_grandiose_names like "Portal", wow! sounds cool! D-BUS.. R-Retard-BUS when?
>>109146362There are even deeper level where linux sucks and that being gpu driver level, in w32 even MS figured out that generic driver model wasn't going to cut it out with gpus and thus wddm was born https://learn.microsoft.com/en-us/windows-hardware/drivers/display/windows-vista-display-driver-model-design-guide , so in short gpu vendors know exactly what their drivers need to do and how to do it. Meanwhile torvalds is screeching how nvidia is hard to work with, without realizing that linux is lacking anything like wddm that tells gpu manufactures how to shoot web https://learn.microsoft.com/en-us/windows-hardware/drivers/display/roadmap-for-developing-drivers-for-the-windows-vista-display-driver-moSo in short, linux is so fucked from deep down desktop wise that it isn't GTK vs QT and not even X11 vs Wayland that's the problem. Meanwhile funnily enough input shenenigans are least problem with linux.
>>109143222>open wm config file>bind ctrl f "musicnigger --play-pause"retards filtered yet again
>>109143498>Yes, exactly this. Commonly used by screen recorders, password managers, etc.Best example are music players. Keep pressing that pause button on your headphones goyim, it won't work because of course who the fuck keeps their music player in focus.
>>109149265Not a portable solution.>>109149341That's indeed another good one.
>>109149837of course its not portable, neither is the X11 solution. what do you expect?
>>109149900X11 solution is portable to everything X11. How many X11 alternatives are there? One, Wayland. WM solution is only for that WM. Now how many WMs are there around, hmm...?It's retarded either way, but don't act like it's even remotely comparable.
>>109149837where you porting it to?
>>109149930>don't act like it's even remotely comparableits literally the same you fucking mongoloid>i3 config: bindsym $mod+r nigger.sh>sway config: bindsym $mod+r nigger.sh
>>109143222Had to work with portals and look into the libportal code at one point. I can confirm that it is trash.>>109143593ywnbaw
>>109149948>>109150017KDE, GNOME, Hyprland, Sway, LxQt, Labwc, et cetera.>>109149930You get it.
>>109150094the X11 solution is exactly the same you ape. you have to configure the keybind in the config of whatever wm you are using. the config for i3 is different to bspwm, which is different to ratpoison, which is different to dwm, etc.
>>109150154In X11, each application can register keybindings on its own with XGrabKey, which works on all X11 desktop environments. Wayland removes this.
>>109150154>X11 program does it for you is same as you fucking around with config files depending on what your distrohopping ass has landed on that day.
>>109143222>because keeping existing programs working is too much of a "burden" on Redhat.X11 is not worked on because it's increasingly hard to work on it, nobody wants to do pointless work like a jew in a camp, it's impossible to implement features in it that are considered old at this point, if you love it so much, you can feel free to maintain and improve it yourself, pro tip - you won't because it's fucking impossible to keep that peace of shit alive. It has kept linux desktop back for decades at this point.
>>109150268>just recreate the entire universe on your own if you don't like our slopFuck off.No one person can do that because everything else is dropping support for X11. The only thing shit attitudes like yours will achieve is people sticking to Windows because it actually works.
>>109148056>There are even deeper level where linux sucks>https://learn.microsoft.com/en-us/windows/win32/learnwin32/overview-of-the-windows-graphics-architecturehere windows starts sucking too..> ..introduced WDDM in Windows Vista (WDDM 1.0) and continues to evolve it with every Windows releaseso maybe they will meet somewhere in the middle and equalize. hard to find HDDs these days, some GPU-accelerated Nvidia drives, hehehe
>>109150268Imagine if instead of wasting effort on their misguided attempt to write a replacement they had dedicated all that energy into cleaning up and fixing X11. Imagine how much nicer X11 would be with 20 years of dedicated effort to improve it... And unlike Wayland, it would actually work (as it does anyway).
>>109148056You're fucking retard and don't know what are you talking about. Linux DRM subsystem is very well documented and there are drivers like Asahi which was made from scratch and working great. If developer have troubles implementing some feature he can always look at the source code for other well-established drivers like Intel or AMD. That's why much there are drivers for much more vendors than on Windows: https://en.wikipedia.org/wiki/Direct_Rendering_Manager#Hardware_supportNvidia working like shit on Linux is because:1. They don't care2. They were developing proprietary drivers and most DRM symbols are GPL protected so they couldn't implement some DRM features3. Instead of cooperating they tried to force their vision on DRM subsystem (without even having upstreamed driver) and were told to go fuck themselves (see GBM and EGLStreams).
>>109143222Usecase for hotkeys?What makes you think consistency is a metric?
>>109150396>most DRM symbols are GPL protected so they couldn't implement some DRM featuresWhat a self-own by Linux.
>>109150416>when you write code for GPL software you must license it under GPL as wellWhat a shocker
>>109143222My application is probably the first real application to implement GlobalShortcuts and my recommendation is to not use it. It broken by design. It works differently in different wayland compositors because of undefined behavior in the protocol definition and because the protocol allows it. It doesn't work properly in any wayland compositor and in some wayland compositors it actually crashes the desktop portal. It works best in kde plasma.
>>109150268Why do wayland retards just make up shit? For example vr support. X11 got vr support designed and implemented in 2 months. In wayland it took 3 years, and then additional years for gnome to support it. This is exactly because of the philosophy design difference between x11 and wayland
>>109150396To be fair, it's only an issue with the regular desktop nvidia gpus. With nvidia jetson they have open source drivers and its proper drm drivers, because those drivers are newer drivers, made from scratch
>>109150823>proper drm driversclown world of GNUnix
>>109150396>That's why much there are drivers for much more vendors than on Windows: https://en.wikipedia.org/wiki/Direct_Rendering_Manager#Hardware_supportmost of those are embedded hardware where windows was never meant, so no point having windows driver, other part is assortment of virtual machine guest drivers, which of there are plenty wddm ones.
>>109152899>most of those are embedded hardwareThese are not for embedded hardware, retard. Mediatek, Adreno drivers are for mobile chips that are powering phones, tablets, pretty much regular end-user hardware.>where windows was never meantWhat kind of argument is that? Next you will tell that Winslop 11 never meant to run without TPM2 and SecureBoot so compatibility with systems before 2018 is irrelevant.DRM subsystem is very powerful and has a lot of standardized features. Last addition of DRM color pipelines made it possible to offload color operations to hardware blocks, which is essential of HDR modesets, lowering TDP, load and latency. These blocks were around for ages but in "great" wddm model there was no support for them and all color operations were performed by shaders like cavemen. And now it's possible to use them if you're using Linux. AMD even develop its own compositor to test their hardware features since on Windows if you're not in bed with M$ (like NVIDIA) you can't do shit.
>>109152988>These are not for embedded hardware, retard. Mediatek, Adreno drivers are for mobile chips that are powering phones, tablets, pretty much regular end-user hardware.Fucking shitass brainlet, mobile phones/retardphones are embedded devices.>What kind of argument is that? Next you will tell that Winslop 11 never meant to run without TPM2 and SecureBoot so compatibility with systems before 2018 is irrelevantFrom begin with Windows was for x86 desktop with server x86 coming along it. And for what parts there was Windows Phone OS, those had wddm drivers just fine from Windows Phone 8.1 onwards. Also there most likely are wddm drivers for most of mobile hardware, those are just sitting in whatever MS research lab that will most likely never produce another Windows Phone OS as embedded environment is plenty different from desktop one starting from very fact that with desktop you want all possible drivers, with embedded you only want certain ones for device internals and planned peripherals. Meanwhile Windows ARM desktop/laptop ventures have been just as retarded as similar linux/android ones.Also fucking funny that you bring up TPM and Secureboot and not any other CPU instruction set requirements that were raised in between windows versions. And yes, OS can wholly be designed for newer hardware and doesn't need to run on older. So whenever MS pulls new major version out of its ass, they can slap whatever hardware age requirements on it they come up with, be those requirements retarded or not. Simply don't buy in propaganda of needing to get new slop and use whatever version was newest for hardware you are using.
>>109152988>DRM subsystem is very powerful and has a lot of standardized features. Last addition of DRM color pipelines made it possible to offload color operations to hardware blocks, which is essential of HDR modesets, lowering TDP, load and latency. These blocks were around for ages but in "great" wddm model there was no support for them and all color operations were performed by shaders like cavemen. And now it's possible to use them if you're using Linux. AMD even develop its own compositor to test their hardware features since on Windows if you're not in bed with M$ (like NVIDIA) you can't do shit.At least read your own fucking documentation about it before spewing shit https://docs.kernel.org/gpu/rfc/color_pipeline.html >How are other OSes solving this problem?>Most OSes will specify the source content format (color gamut, encoding transfer function, and other metadata, such as max and average light levels) to a driver. Drivers will then program their fixed-function HW accordingly to map from a source content buffer’s space to a display’s space.>When fixed-function HW is not available the compositor will assemble a shader to ask the GPU to perform the transformation from the source content format to the display’s format.>Why is Linux different?>Unlike other OSes, where there is one compositor for one or more drivers, on Linux we have a many-to-many relationship. Many compositors; many drivers. In addition each compositor vendor or community has their own view of how color management should be done. This is what makes Linux so beautiful.>This means that a HW vendor can now no longer tune their driver to one compositor, as tuning it to one could make it look fairly different from another compositor’s color mapping.So whole shit is just thing that exists because its linux and no such thing is needed in windows / macos
>>109143593>Starts talking about Qt/GTK>In a conversation about Wayland being shitYou're a fucking retarded vibe coder. Go coom to lolis instead of shitting up this board. Nigger.
>>109153447>Fucking shitass brainlet, mobile phones/retardphones are embedded devices.Retard https://en.wikipedia.org/wiki/Embedded_system>those had wddm drivers just fine from Windows Phone 8.1 onwardsWindows Phones never used wddm drivers, wddm is only for desktop>Also there most likely are wddm drivers for most of mobile hardware, those are just sitting in whatever MS research labAt this point you're just making shit up?>Also fucking funny that you bring up TPM and Secureboot and not any other CPU instruction set requirements that were raised in between windows versions. And yes, OS can wholly be designed for newer hardware and doesn't need to run on older. So whenever MS pulls new major version out of its ass, they can slap whatever hardware age requirements on it they come up withThat was not the point. Point was your retarded assumption that the reason there are so little display drivers on Windows compared to Linux is that "windows wasn't designed for such devices". That's bullshit and has nothing to do with reality. Real reason is that Windows is a closed source shitware where you can't develop a driver unless you're in close contact with MS, not to mention to introduce a new feature in graphic subsystem. Meanwhile on Linux everything is open source, all documentation and most features are vendor-agnostic, meaning that literally who can develop a fully-featured GPU driver without permission from Linux Torvalds himself (and yeah, if you managed somehow to make your own driver on Windows you also have to buy signing key from M$ partners if you want for end-user to install it without entering debug mode). And everyone can participate in development and send patches to kernel (that what KWin developers are doing, and AMD devs are in turn send patches to KWin). Meanwhile, on Windows you have shit like this and no-one can do anything about it: https://github.com/dylanraga/win11hdr-srgb-to-gamma2.2-icm
>>109153466>At least read your own fucking documentation about it before spewing shitYeah, yeah, sure thinghttps://learn.microsoft.com/en-us/windows/win32/wcs/display-calibration-mhc#new-gpu-color-transform-pipeline>The pipeline isn't directly programmable by apps, and instead is exposed only via MHC profiles; see below for more details.So it's only used when you have specific color profile and to convert compositor color buffer to display's one, while compositor still using pixel shader to convert content color space to its own.
>>109143593>dbus is fine because you're forced to use itRed Hat/Gnome logic in a nutshell
>>109153713>Retard https://en.wikipedia.org/wiki/Embedded_system>Telecommunications systems employ numerous embedded systems from telephone switches for the network to cell phones at the end user. Computer networking uses dedicated routers and network bridges to route data.>Consumer electronics include MP3 players, television sets, mobile phones, video game consoles, digital cameras, GPS receivers, and printers. Household appliances, such as microwave ovens, washing machines and dishwashers, include embedded systems to provide flexibility, efficiency and features. Advanced heating, ventilation, and air conditioning (HVAC) systems use networked thermostats to more accurately and efficiently control temperature that can change by time of day and season. Home automation uses wired and wireless networking that can be used to control lights, climate, security, audio/visual, surveillance, etc., all of which use embedded devices for sensing and controlling.As i said, fucking shitass brainlet.>>109153721>So it's only used when you have specific color profile and to convert compositor color buffer to display's one, while compositor still using pixel shader to convert content color space to its own.And thus for example my DX12 projects use FP16 backbuffer so there isn't useless extra conversion between program and dwm, so its thing programmer can choose to optimize for or not. Guess what happens in linux between program and compositor when they have different buffers.
>>109153777>And thus for example my DX12 projects use FP16 backbuffer so there isn't useless extra conversion between program and dwm, so its thing programmer can choose to optimize for or not.How is this related to HDR? On KWin fullscreen applications bypassing compositors and tone mapping since it's all getting offloaded to driver. On Windows you always have tonemapping in the form of pixel shader because that's how dwm works.
>>109143222global shortcuts = keylogger
>>109152988>essential of HDR modesetswow, an abbreviated nothingburger! GNUnixoids are real clowns
>>109153806>How is this related to HDR? On KWin fullscreen applications bypassing compositors and tone mapping since it's all getting offloaded to driver. On Windows you always have tonemapping in the form of pixel shader because that's how dwm works.There are similar optimizations in windows, but they fully apply only if there are no extra overlays, as if there are overlays, then dwm jumps back into loop to blend shit together. Then other is that do you match display output format or do you match dwm format, that also mainly already decides what color space your data portrays in that byte format. https://devblogs.microsoft.com/directx/demystifying-full-screen-optimizations/https://learn.microsoft.com/en-us/windows/win32/direct3darticles/high-dynamic-range
>>109153466>https://docs.kernel.org/gpu/rfc/color_pipeline.html> What problem are we solving?the "problem": GPU in a monitor does some "hardware acceleration" wespeak about as good and have to have to support.
>>109153713>Windows Phones never used wddm drivers, wddm is only for desktophttps://learn.microsoft.com/en-us/windows-hardware/drivers/develop/driver-model-convergence>Display Driver Model (WDDM) 1.3 runs on Windows 8.1 and Windows Phone 8.1.
>>109153713>Meanwhile, on Windows you have shit like this and no-one can do anything about it: https://github.com/dylanraga/win11hdr-srgb-to-gamma2.2-icmThis is for game developer to handle and Windows gives more than enough tools in DirectX for it, so if game doesn't give you control over gamma / other color options, then that was decision of game programmer not to bother with it.
>>109153808You can make a global shortcuts API that only triggers when modifiers are pressed, and then doesn't send the intercepted hotkey to any other program. It'd be impossible to make a keylogger that way.https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-registerhotkeyThat'd be how a wayland protocol would do it.
>>109143593dbus is shit because what's actually useful is IPC but instead you get RPC and dumb OOP shit
>>109150331>imagine if someone just solved impossible problems, wouldn't that be nice?>>109150794What about the HDR support?>>109150319That's what wayland devs did, they recreated the entire universe because nobody liked your slop.
>>109156414>muh hdrThis is such a non-issue, literally no one cares about this.>they recreated the entire universe because nobody liked your slop.And they did a horrible job of it.
>>109157033There are more people with HDR capable monitors than people with VR headsets
>>109158041K. No one cares about VR either.
>>109158044>No one cares about VR either.how to rise and launder OSS fund money without PBRMKR VR AMR VKDRM etc features then? refactoring is a puff nobody cares about. new features is an essencial holy cow to milk.--im glancing those MS manuals, they are kinda mr.Teacher says good and teaches you good, while GNUnixoid wespeak manuscripts mention some kind of "problem solving" and attempts of justification.
WHAT'S THE USE CASE FOR GOOEY?
>>109143222I had a one-line config file under X11 that let me press ctrl-space to clickWayland made me install a gay ass GUl attached to a systemd service and dbus to do the same thing for no reason. I couldn't even install the thing as a user service because it needed special permissions anywayIt's baffling how bad an idea "bring your own compositor" has been
>>109158233this but unironicallydon't we all use text buffer gooeys in emacs anyways?
>>109158251lmaoI wonder what the non-systemd version of that would've been.
