https://bugs.kde.org/show_bug.cgi?id=485034

            Bug ID: 485034
           Summary: Feature requests for Custom Tiling
    Classification: Plasma
           Product: kwin
           Version: 6.0.3
          Platform: Other
                OS: Linux
            Status: REPORTED
          Severity: normal
          Priority: NOR
         Component: Custom Tiling
          Assignee: kwin-bugs-n...@kde.org
          Reporter: tony0000...@gmail.com
                CC: notm...@gmail.com
  Target Milestone: ---

So, all of this actually was a not so laid down "monologue" in Niccolò
Veggero's Telegram group. I am just reporting it here so that we all can
discuss it.

As a tiling scripts user, I feel as though the Tiling API is incomplete and
this is painfully evident when, pre-Tiling API, Bismuth would work really well
but, post-Tiling API, Polonium (which uses the Tiling API) is very much
limited.

In which ways, you ask?

The first enormous issue is that windows have to be shift-dragged to snap to
tiles. This is okay if we're looking at tiles like they are implemented in
PowerToys and it does overlap largely with that scope, but it isn't how tiling
window managers really work. In a tiling environment windows should be able to
be just dragged to other tiles, and the layout should be able to accomodate for
the change (sure, this is up to the script that is providing the actual tiling
engine implementation, but still it cannot do this now and so it pains a lot of
people.)

Talking about shift-dragging, we should be able to shift-drag floating windows
even if we only have one big tile. Currently KWin won't allow shift-dragging
anywhere if you only have one window tiled in one tile, but Polonium exploits
the shift-drag functionality to split that tile into two and place the other
window besides the already placed one.

Lastly, there should be a way for tiling scripts to be able to provide clear
highlighted insertion points for windows (which would require one less hack -
that is, what I described above.) This could be done by repurposing desktop
corner snapping, or by providing edge-snapping inside tiles.

And still, I feel as though this would only scrape the surface. KWin's Tiling
API is a big promise, but far from complete; yet, I hope script developers and
KDE contributors can collaborate towards a common goal here.

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to