> GTK4 goes more in the direction to alienate everything that isn't GNOME. The opposite of what Qt does. People might have different experiences, but the fact that QT **still**, even in version 6, does not support `text_input_unstable_v3` - after over five years - and thus doesn't have onscreen keyboard support on Gnome or Sway speaks to the opposite, IMO. In this particular case it instead implements some downstream alternatives (v2 and v4)[https://github.com/qt/qtwayland/blob/5cdc706747feca4e62928332350f0229a9788ce8/src/client/qwaylanddisplay_p.h#L50-L52], v2 being supported by Kwin - but nothing of that is upstream in wayland-protocols. So even if we were to port to the latest QT version OSK would only work on Kwin (and maybe some niche compositors). That being said, I personally would favor freezing X11 on GTK3 and only move the Wayland bits to something new - be it another toolkit or a full own implementation (which IMO could well end up smaller/lighter than the glue code to any existing toolkit - *maybe* apart from theming support, if we want to have that).
Bug 1701123 Comment 40 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
> GTK4 goes more in the direction to alienate everything that isn't GNOME. The opposite of what Qt does. People might have different experiences, but the fact that QT **still**, even in version 6, does not support `text_input_unstable_v3` - after over five years - and thus doesn't have onscreen keyboard support on Gnome or Sway speaks to the opposite, IMO. In this particular case it instead implements some downstream alternatives [v2 and v4](https://github.com/qt/qtwayland/blob/5cdc706747feca4e62928332350f0229a9788ce8/src/client/qwaylanddisplay_p.h#L50-L52), v2 being supported by Kwin - but nothing of that is upstream in wayland-protocols. So even if we were to port to the latest QT version OSK would only work on Kwin (and maybe some niche compositors). That being said, I personally would favor freezing X11 on GTK3 and only move the Wayland bits to something new - be it another toolkit or a full own implementation (which IMO could well end up smaller/lighter than the glue code to any existing toolkit - *maybe* apart from theming support, if we want to have that).