But you just want to shit of GTK for no reason. It's an open source project, so they owe you nothing, and nothing was stopping you from contributing your own improvements.
The first instance of the bug report happened in 2014 when mclassen introduced it and refused to fix it. The most recent instance of the bug report happened 7 months ago when I proved to them they'd ported the bug from gtk3 to gtk4 too: https://gitlab.gnome.org/GNOME/gtk/-/issues/5872 I even showed them a partial patch for the problem for gtkfilechooserwidget.c to restore default text entry input but the filechooser is such spaghetti that everyone is afraid to change anything.
So, yeah, the bug still exists. And it really is more important than another renderer. This is basic functionality that's been missing for a decade.
they owe you nothing, and nothing was stopping you from contributing your own improvements
lands inside the [ignorance..rudeness..hypocrisy] triangle most of the times. It would be more agreeable if stated as:
they owe you nothing, and nothing was stopping you from contributing approved improvements from their backlog
The same bug exists in gtk4 too and they also refuse to fix it there. https://gitlab.gnome.org/GNOME/gtk/-/issues/5872
So, to conclude, everything you said is wrong re: this specific instance even if it may apply generally.
Do they differ in this respect from "Modern big proprietary projects" ?
If this is a characteristic of "modern big software projects" then it has nothing specifically to do with whatever happens or doesn't happen in open source development.
The "nothing" stopping me from contributing is that GNOME source is in an inconvenient location, is more difficult to understand coming from non-GUI microcontroller programming than for the core devs, and is a hassle to build. When I try to install or compile dependencies, I seem to encounter a new error every time, like GNOME Builder not respecting a corporate proxy while using WSL even though my environmental variables are set correctly. (So now, you're learning about "curl -vvv" and maybe installing MSYS instead of figuring out why your mouse doesn't scroll up. Make sense yet?)
Beyond that, I don't want to turn my computer into a machine with Flatpak and multiple GB of devel files just so I can fix the scroll bar that should already work and might be caused by libinput or kernel source.
Something is fundamentally wrong if the GTK scrollbar---one of just a handful of available widgets that would be put in countless GTK programs---doesn't work in a GNOME app.
The last 3 Linux devices I've owned have been an overpriced garbage heap of Certified hardware---running too hot with GPU glitches, OS hanging, everyone's special take on application packaging (aka No More Ubuntu for me), awful sleep and power management, terrible peripheral support, having to learn about journalctl and a laundry list of kernel parameters...
The thing in common with those terrible modern devices that weren't around on my much-loved XPS 13 9370?
Wayland
GTK 4
Pipewire
S0ix
Kernel 6+
I'm not necessarily blaming one specific new technology, but I just today took a video of my laptop (2500 euros, shipped with linux by a major manufacturer) evidently having graphics glitches just from opening Firefox (on a Radeon, just to clear up that not every GPU Linux problem is Nvidia and everyone using AMD is enamored). I have no idea if it's a software or hardware problem---it also happens if I change the UI scale to 125% or plug in a dock---but it's not even the first of ten major problems this laptop has had.
I'd gladly RMA the thing, but then I don't know what to develop on except maybe an Android tablet cross-compiling to x86 plugged into a dock.
My grandpa's 286 was more reliable than an off-the-shelf Linux or Windows laptop today, and the only MBP I ever used left a sour taste.
Perhaps but your grandpa's 286 couldn't have animated emojis made out of gamma correct antialiased vectors in button caption for its GUI :-P
You mean that the GTK/GNOME people declare they will never fix their file chooser? I've pretty much figured that out. That in itself is more of a problem than the state of the file chooser, because fixing faulty design is one thing, fixing but dogmatic insistence on poor UI is another thing altogether.
> But you just want to shit of GTK for no reason.
I disparage the GTK file chooser because it's horrible to use. I don't know anybody who works on GTK. I don't write software which competes with GTK (I'm a GPU guy). I have better things to do with my time than bad-mouth software projects I have no personal interests for or against.
What's actually happening is that some peopl are holding your hands over your eyes and refusing to see that (some of the) GTK UI elements are utterly broken intentionally and for years or decades. And while GNOME apps have alternatives, GTK is, unfortunately, popular among apps which don't really have alternatives, so I'm stuck with it.
> It's an open source project, so they owe you nothing
Huh? It's a project of human society to serve people's needs. So they owe users and developers to do a decent job and address those needs. You seem to be suggesting that if I don't personally pay them then I should just shut up and accept their choices.
I would never ever say something like that to a user of my FOSS - and I do owe my users, a lot.
> and nothing was stopping you from contributing your own improvements.
1. _Everything_ is stopping me from contributing an improvement. The idea is rejected on principle.
2. I understand you are offering to volunteer take up maintenance of the FOSS work I'm doing while I go off and start getting into GTK, with which I have no experience. Or perhaps you want to take over my day job so I can have more spare time?