search

probonopd / linuxdeployqt

Makes Linux applications self-contained by copying in the libraries and plugins that the application uses, and optionally generates an AppImage. Can be used for Qt and other applications
Other
2.22k stars 414 forks source link

Add support for the wayland platform #540

Open buresu opened 2 years ago

buresu commented 2 years ago

The following library and plugins are required to support the wayland platform.

libQt5WaylandClient.so
libQt5WaylandCompositor.so
platforms/libqwayland-*
wayland-*

close #189

ardera commented 1 year ago

@probonopd Friendly ping, do you have any time to look at this? :)

probonopd commented 1 year ago

Thanks, but I am not interested in Wayland. None of my systems run Wayland, because it never works quite right in my experience. Especially when you are not using GNOME or KDE.

tulpenkiste commented 1 year ago

Thanks, but I am not interested in Wayland. None of my systems run Wayland, because it never works quite right in my experience. Especially when you are not using GNOME or KDE.

Hi probonopd, I do not understand the reasoning you are using here to continue ignoring Wayland support on your software.

When it comes to not being interested in Wayland, that is fine. But some people (like myself) use Wayland on a daily basis (even using it exclusively).

However, this PR that you've rejected would have not affected you in the slightest as XOrg support would still be usable. It would have just improved the experience on Wayland as we wouldn't need to rely on xwayland to even be able to run it.

Plus, when it comes to Wayland "not working right", I can say the exact same about XOrg as it had always been janky and unstable for me (including but not limited to random crashes, random screen tearing and widgets acting weirdly in KDE Plasma) which I cannot say the same about Wayland. Even then, if I do make an application that uses a graphical interface, I won't just skip XOrg support because that would be quite odd. So why should it be the same for Wayland support, especially when the PR gave you literally all the information you needed to implement it?

These kinds of decisions by you on AppImage related projects makes me consider things like AppImages last when going to obtain new software. Right behind compiling software from source and Flatpak.

mothdotmonster commented 1 year ago

I think that it's a good idea to add support, as part of the reason why Wayland isn't quite ready for prime time is because not everything supports it yet! So adding extra support to things will make Wayland better.

probonopd commented 1 year ago

So adding extra support to things will make Wayland better.

Thanks, but I am not interested in making Wayland better. I think that is the job of the people who are pushing Wayland - to make it run existing X11 applications flawless without requiring changes in the applications.

It has been claimed that Wayland can run X11 applications using XWayland. So it should be able to run "normal" applications just fine?

probonopd commented 1 year ago

Code change looks reasonable; however we need to find out whether this causes any issues for non-Wayland users, and whether this draws in additional dependencies which increase the AppImage size significantly. Any other potential risks/downsides?

Since I don't run Wayland I can't test the Wayland side of this.

Remains the philosophical question: Should we really help promoting the splintering of the *nix desktop by shipping Wayland-specific stuff? I tend to think that the better way is to make AppImages use X11 (since that is what most systems use) and have Linux distributions who insist on already shipping Wayland ensure that Wayland can run X11 applications very, very well.