GTK 4.0 Isn't Expected To Be Released Until Autumn 2020

[phoronix]

Phoronix: GTK 4.0 Isn't Expected To Be Released Until Autumn 2020

GTK 4.0 won't be out this year, nor is it expected next spring as part of the GNOME 3.36 cycle, but now the developers believe this next major tool-kit update will be ready to ship in just over one year's time with the autumn release of GNOME 3.38...

http://www.phoronix.com/scan.php?page=news_item&px=GTK4-Features-Delay-Autumn-2020

[Guest]

I expected the GTK team to drop theming from GTK 4. Considering my past experiences, I'm trying to steer away from software that uses their awful toolkit, as I've had the misfortune of both using and writing GTK applications.

[Daktyl198]

I hate these days where there always has to be two versions of every app. GTK3 apps work like hell on non-Gnome systems, so somebody has to make a Qt version, and while the Qt version of an app works fine on Gnome, it doesn’t “fit the look” so somebody makes a GTK version.

wish GTK would work more toward cross compatibility with non-gnome systems if so many developers are dead-set on using such a broken toolkit.

[ElectricPrism]

I expected the GTK team to drop theming from GTK 4. Considering my past experiences, I'm trying to steer away from software that uses their awful toolkit, as I've had the misfortune of both using and writing GTK applications.

That would be autistic, the only selling point on GTK over other offerings is how customizable it is. And just because an implementation has problems doesn't mean a vision is wrong.

Changing vision on a whim every major version number is what made Gnome 3.0 and KDE 4.0 so fucked up and unusable for those who were around to remember the sudden workflow fuck moving from Gnome 2.x => 3.0 or KDE 3.X => 4.0

Changing vision and core mechanics is the LAST thing I want, what could they possibly offer us that we don't already have? Transparent Gaussian Blur Backgrounds just to compete with KDE?

Proper Global Menu tie ins via D-Bus for searchable file menu HUDs and Global Menu extensions?

Open source deserves a comic strip about Point OH releases since the terms Alpha and Beta don't seem to be in our vocabulary. Point OH on commercial software means it'll work, on Open Source it means you're in for a could-be lethal safari ride.

[BraindeadBZH]

GTK is bound to lag far behind Qt from now on. I don't see any good reason to choose GTK over Qt for development.

[Britoid]

I hate these days where there always has to be two versions of every app. GTK3 apps work like hell on non-Gnome systems, so somebody has to make a Qt version, and while the Qt version of an app works fine on Gnome, it doesn’t “fit the look” so somebody makes a GTK version.

wish GTK would work more toward cross compatibility with non-gnome systems if so many developers are dead-set on using such a broken toolkit.

Elementary and Purism are non-GNOME systems that use GTK.

Qt has 10x the codebase and 10x the engineers that work on it compared to GTK, and was made from the beginning to be able to pretend or look native on multiple systems. It's a monster of a toolkit and it's unfair to expect a toolkit with 1-2 developers to be able to have the same feature set.

[Britoid]

That would be autistic, the only selling point on GTK over other offerings is how customizable it is. And just because an implementation has problems doesn't mean a vision is wrong.

Changing vision on a whim every major version number is what made Gnome 3.0 and KDE 4.0 so fucked up and unusable for those who were around to remember the sudden workflow fuck moving from Gnome 2.x => 3.0 or KDE 3.X => 4.0

Changing vision and core mechanics is the LAST thing I want, what could they possibly offer us that we don't already have? Transparent Gaussian Blur Backgrounds just to compete with KDE?

Proper Global Menu tie ins via D-Bus for searchable file menu HUDs and Global Menu extensions?

Open source deserves a comic strip about Point OH releases since the terms Alpha and Beta don't seem to be in our vocabulary. Point OH on commercial software means it'll work, on Open Source it means you're in for a could-be lethal safari ride.

I thought exporting a menu over dbus was supported by GMenu for accessibility services, so gtk modules were no longer required (and they shouldn't, they're ugly).

GTK is bound to lag far behind Qt from now on. I don't see any good reason to choose GTK over Qt for development.

C++..

[Shaheed]

C++..

Let me FTFY:

C++ or Python.

[BraindeadBZH]

C++..

Qt makes a good job at "abstracting" C++ through a very clean API, also now, a lot of code is written in QML/Javascript within Qt applications.

[kpedersen]

also now, a lot of code is written in QML/Javascript within Qt applications.

This is not a good thing.

From the QML/Javascript; imagine that you need to call into... i.e Steam's C++ matchmaking API. What would you do? Yep; spend your life writing bindings rather than completing the actual project and releasing your project.

Same goes for the stupid Qt MOC build system and qmake; integrating with existing CMake projects (most industry C/C++ libraries use this) means you are spending your time masturbating Qt rather than creating solutions and making money.

Yes; at least you are up to that point. With Gtk4, you are still trying to get the sprawling cancerous mess building on Win32 (for the biggest market).

UI with C++ is a little bit of a joke. For most projects I use wxWidgets (Gtk on POSIX, Win32 on Windows) but there are issues; it loves unsafe raw pointers for one.