[Crinus]

> The way I see it is that a new toolkit like GTK or Qt is required in the target language.

The way i See it is that a new toolkit like GTK or Qt is required in C so that it can interface with any language as to avoid unnecessary duplication between languages (it is much easier to convince people switch toolkit than switch programming languages), ensure no language-specific indirect dependencies (especially if it is to included in OSes like OpenBSD that are traditionally C oriented) and with a stable backwards compatible ABI to avoid wasting the time and effort of whoever maintains the bindings (no need to rewrite them every few years), whoever decides to learn the new API (no need to learn something new for doing the exact same thing as before, only now in a slightly different yet incompatible way) and perhaps hopefully become a "standard" API people can target on Linux without having to carry 29892482923TB of libraries for a trivial app since you cannot rely on anything outside of X11 being there (and if GNOME people have their way, you wont be able to rely on even that). And of course all that in a permissive license that leaves everyone happy.

Now guess how likely all of this is going to happen.

[jcelerier]

> and perhaps hopefully become a "standard" API people can target on Linux without having to carry 29892482923TB of libraries for a trivial app since you cannot rely on anything outside of X11 being there

so, uh, what would you depend on ? at the bare minimum you would need, for a gui toolkit :

- dependencies to X11 (xcb) and wayland (and cocoa and win32)

- dependencies to xkb, etc.. for keyboard handling

- dependencies on ICU for unicode text handling

- dependencies on freetype, fontconfig, harfbuzz, pango, etc... for font rendering

- dependencies on GL to allow your users to have a GL viewport, and potentiall on a software implementation of GL on windows because yes, even in 2019 this is still a problem on some user's computers

- dependencies on cairo, skia, QPainter, or something like that to provide a canvas-like API

- dependencies on libinput because that's the modern way to get mouse input on linux

- you are going to want a library of useful datastructures that aren't provided in C -> dependency to GLib or something else

- dependencies on libpng,jpeg, etc.... because you want your users to be able to have images in their app

- dependencies on dbus to be able to provide desktop notifications, etc

and you aren't even starting to match all the useful functionalities of GTK & Qt

[Crinus]

Of everything that list all you really need is the X libraries and you do not need to carry (meaning distribute) those since they are part of the every single desktop environment and have a stable ABI. The rest aren't strictly necessary, but i think Cairo has a stable ABI. I'm not sure if it is available everywhere though. OpenGL (and Vulkan) can be provided via dynamic linking, though only via an API that doesn't guarantee its existence (it doesn't exist in all X11 desktops - AFAIK Raspberry Pi doesn't support them).

In any case, the dependencies i mean here are dependencies for the applications that target your API, not the dependencies your API has. You can make your API use whatever library you want to -e.g.- load PNG, JPG or whatever, but that can be hidden from the user of the API (ie. their binary will link against libyourapi.so.1 for yapi_load_image and what that specific version of libyourapi.so.1 uses to load the requested image is an implementation detail).

Note that this is in the context of making that API a standard people can target, not a reusable library (and its own dependencies) people carry with their binaries. Basically the equivalent of how you do not need to give USER32.DLL on Windows to your users because you can pretty much expect that it will be there.

Nowadays the only equivalent on Linux is the X libraries, anything else is either not guaranteed to be there now, or not guaranteed to be there in ~5 years because thanks to CADT people cannot stick with an API and feel the need to break everyone's programs to feel useful.

[jcelerier]

> Of everything that list all you really need is the X libraries and you do not need to carry (meaning distribute) those since they are part of the every single desktop environment and have a stable ABI.

both Debian and Fedora use wayland by default now. And wayland alone only gives you the ability to open a viewport, so you are going to need a lot of stuff if you want your UI library to do something as simple as :

    #include <linux_ui.h>
    int main() {
        linux_window* window = linux_ui_create_window();
        linux_widget* label = linux_ui_create_label(u8"お早う御座います");
        linux_ui_widget_set_parent(label, window);
        linux_ui_widget_set_pos(label, 50, 50);

        return linux_ui_process_events();
    }

> Note that this is in the context of making that API a standard people can target, not a reusable library (and its own dependencies) people carry with their binaries. Basically the equivalent of how you do not need to give USER32.DLL on Windows to your users because you can pretty much expect that it will be there.

But almost no one makes Win32 apps with the Win32 API anymore anyways. The majority of desktop windows apps created today are using WPF or UWP which does not use any "native" stuff (that is, the win32 UI API), Qt or Electron. If this model does not even work on windows, why would you want to port it on linux instead of letting people ship their app as they wish ?

> Nowadays the only equivalent on Linux is the X libraries, anything else is either not guaranteed to be there now, or not guaranteed to be there in ~5 years because thanks to CADT people cannot stick with an API and feel the need to break everyone's programs to feel useful.

Even if someone made this magical new library today, I would still have to package it along with all its dependencies in an AppImage because I have users still on ubuntu 12.04, 14.04, and 16.04 as well as debian stretch and jessie.