hw posted on Tue, 20 Feb 2024 17:54:27 +0100 as excerpted:
> Why does this happen[1], how can I prevent it from happening, and how do
> I get the lost window back?
It looks like you already got an answer to this so I'll skip it. (The
post is long enough...)
> [1]: I think I moved it into the top left corner to see if it would be
> maximised and/or get tiled and then it somehow disappeared. In System
> Settings-->Workspace Behaviour-->Screen Edges,
FWIW, different path in plasma6. (plasma) system settings > input & output
> mouse & touchpad > screen edges.
> both options were enabled
> at the same time by default. That doesn't seem to make sense,
> and I turned the Maximizing off.
I believe you're misunderstanding the configuration. It does make sense,
and here's how:
The screen edges configuration consists of several different areas with
settings that are conceptually separate:
At the top there's an image of a monitor with eight different dots, each
of which corresponds to a corner or edge, with each one of the eight
available to set separately. Edges/corners that have an action assigned
have their dot filled in, while those with no action assigned have their
dot empty.
If you click on one of the eight dots it should give you a list of the
actions you can choose to enable for that edge/corner. Here, you are
correct -- only one can be enabled at once, tho there's a list of many
actions to choose (including "none", thus disabling action triggers for
that corner/edge).
Below that are three checkboxes with the choices that seem to be causing
the confusion here. The fact that checkboxes (as opposed to radio-
buttons) are used here indicates that the options are independently
togglable and that it DOES make sense to have more than one turned on at a
time.
Specifically the three, maybe described a bit more clearly here (6.x
version) than it 5.x?:
* Maximize: [checkbox] Windows dragged to the top edge
Keep in mind that this is a separate behavior from the top edge dot in the
first section. The dot is the behavior when the pointer bashes the top
edge when *NOT* dragging a window. This checkbox would be the behavior
when dragging a window.
Basically, if checked, it just gives you another way to maximize a window,
that some people find easier/faster/more-natural than clicking the
maximize button in the titlebar, choosing it from the window/system
actions menu, hitting the maximize hotkey you may have assigned, or...
But some people will find it annoying, particularly if they drag windows
around a lot, as if they drag it too close to the top of the screen they
might trigger maximize accidentally, when they just wanted to reposition
their window (or take it to a different desktop, see below). So it's an
option.
* Tile: [checkbox] Windows dragged to the left or right edge
Again, this behavior is separate from the left/right edge trigger
configurable from the appropriate dot above. Again, that would be when
not dragging a window, while this checkbox controls what happens if you're
dragging a window when you bash that left or right edge. And just as top
is not side, this checkbox toggles tiling, not maximizing.
And again, some people will find this annoying and that it triggers
accidentally for them when they were just trying to reposition a window
(or trying to drag the window to a different desktop, see below).
And just to stress by repetition. Yes, it does make sense to have drag-
to-top-to-maximize and drag-to-side-to-tile as separate independently
togglable actions, since people might want one or both or none, and
they're to entirely different edges. Thus the checkboxes, indicating
independently configurable toggles.
* Behavior: [checkbox] Remain active when windows are fullscreen
This one we'll discuss below your question about it, quoted below...
> BTW, what is the behavior 'Remain active when windows are fullscreen'
> supposed to do? Deactivating windows once they're fullscreen doesn't
> make sense. What is remaining active, or does not remain active? If
> it's a behaviour that remains active or not, then I can only say that
> activating or deactivating an unknown behaviour isn't exactly helpful.
What this checkbox does is toggle whether the EDGE TRIGGERS remain active,
not the WINDOWS. If this is disabled bashing the pointer against the edge
(or corner) when a full-screen window is active will do nothing. If it's
enabled the configured action will still occur when you bash the pointer
against the edge/corner.
Might as well complete the discussion, describing the four remaining
settings, below the three checkboxes.
* Trigger quarter-tiling in: [spinbox %] of the screen
This controls how big the "center edge" zone is when dragging-to-tile to
the left or right, in which the tiling is half-max, vs. the upper and
lower quarter-max zones. So (I think) it's only effective when the tile
checkbox is checked.
* Switch desktop on edge: [dropdown listbox]
The choices are disabled, always-enabled, or only when moving windows.
Obviously it doesn't matter if you have only a single virtual desktop, but
for those with more than one it makes a difference, and given the other
available edge-actions and the fact that hotkeys can be configured to
switch desktops and send windows to other desktops, it can make a lot of
sense to disable this entirely here to prevent accidentally switching
desktops.
Alternatively, one may want to enable edge-triggered desktop-switching,
but disable some of the other settings. Either way, in my experience,
having this enabled AND the edge/corner actions in the top section AND the
drag-top/side-max/tile, is simply too much as they conflict. So pick the
set you prefer and disable the others. =:^)
The last two settings are:
* Activation delay: [spinbox ms]
* Reactivation delay: [spinbox ms]
Adjusting these can help prevent accidental triggering. Unfortunately,
it's only the single pair of delay settings that apply to the entire set.
It'd be nice to have separate delay settings for switch-desktop (with
window-drag or not) vs the others (which are naturally differentiated by
whether you're dragging a window or not), so one could for instance real-
short delay = do-nothing, medium-delay (400 ms) = action/tile/maximize,
much-longer-delay (800 ms) = switch-desktop. Kinda like short-tap vs.
long-tap on a phone, or single vs. double-click (which of course is delay-
based as well).
--
Duncan - List replies preferred. No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master." Richard Stallman
