Default touch gesture assignments should be methodical
(Moved from BUG: 478177 as @ngraham suggested that Invent is better suited for this kind of discussion.)
At the time of writing, virtual desktop (VD) switching gestures use 3 fingers. We support VD grids, so there are three-finger gestures allocated for up, down, left and right. Overview uses 4 fingers (up/down) because those were still available. Those seem to be the only ones enabled by default for Plasma 6, based on a quick test. 5 fingers seems to do the same as 4 fingers.
In my opinion, this assignment seems haphazard and will make it harder to add more sensible gestures later. Consider this forum post for an example.
Sure we'll eventually have a way to configure gesture assignments, but we should also think about what makes sense for a roughly intuitive default behavior, so we don't have to change it drastically after Plasma 6.0 or be stuck with something sub-optimal for continuity reasons.
It doesn't have to be my suggested logic, but either way we should have a mental model of "why is this functionality assigned to this gesture" (by default, at least).
Here's my rough suggestion:
- 2-finger gestures should be focused on doing stuff within an app (obviously we can't change this)
- 3-finger gestures should be focused on actions within the current VD
- 4-finger gestures should be focused on actions spanning different VDs (the most zoomed-out level of gestures)
By that logic, I'd like to suggest a change in default gesture assignments:
- VD switching to use 4-finger gestures (up, down, left, right)
- Overview to use 3-finger pinch (to open / zoom in) and spread (to close / zoom back out).
- 4-finger pinch/spread can go directly to the VD switcher view within Overview
- 3-finger up/down/left/right is not assigned by default, reserved for future gestures such as quick window switching
But really I'd be happy with anything that's backed by some kind of logic/methodology to make things easier to memorize.