Summary:
Right now, when there are no configured wallpaper slideshow paths or wallpapers in the
slideshow, the list and grid (respectively) are just empty. This patch adds some
placeholder labels to let the user know that it's not broken, just like we do in the
Thunderbolt KCM.
See one `FIXME:` in the code that I could use a hand with.
Test Plan: {F7870947}
Reviewers: #vdg, #plasma, mart
Reviewed By: #vdg, #plasma, mart
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26528
Summary:
BUG: 381551
FIXED-IN: 5.18.0
Test Plan:
Before: {F7851382, size=full}
After: {F7851383, size=full}
Texted extensively because this code is so brittle. No regressions with long date in a
thick or normal-sized horizontal panel or short date in a thick or normal-sized
horizontal or vertical panel.
The one issue I could find is that sometimes switching between short and long date
formats in a double-thickness vertical panel causes the long date to be truncated again
until plasmashell is restarted. I was unable to determine why. Assistance would be
appreciated if this is a blocker.
Reviewers: #plasma, #vdg, ndavis
Reviewed By: #vdg, ndavis
Subscribers: ndavis, plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26312
Summary:
For normal distros it doesn't make sense, KDE_INSTALL_FULL_DATAROOTDIR
will be /usr/.
For devs using a prefix startplasma-dev.sh already does this for us only
for the dev setup.
For any other user, it should be their responsibility to adjust it.
It's currently broken anyway, if a user simply installs flatpak that sets this env
earlier in the boot, so this will do nothing. It implies no-oe relies on thisl
It's a relic from 2007 (bd38d59e11548f160efd0918ca5461536ac72349) that doesn't
make sense now.
BUG: 315452
Test Plan: Compiles
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26537
Summary:
System tray dialog/popup can be tiled. This should not be allowed, in this case expanded representaiton should have fixed size. When tiled, it is no longer possible to restore original size.
This also applies to other plasmoids.
BUG: 386551
FIXED-IN: 5.18.0
Test Plan:
1. Open any applet from system tray (for example Clipboard or hidden items)
2. Use shortcut to tile window (by default Meta + Left/Righ/...)
3. Applet will change size and tile at the borded of the screen
4. Expected:
a) before fix: allowed and not possible to restore size
b) after fix: not allowed or restores size on next open
Reviewers: #plasma_workspaces, #plasma, davidedmundson, broulik, ngraham
Reviewed By: ngraham
Subscribers: ngraham, plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26545
Summary:
For each tray icon XEmbedSNIProxy is creating container window 32x32 in size. It is black with opaque set to 0 (fully transparent when compositor is enabled). All of these container windows are stacked below all windows, so normally you can't see them. On creation all container windows are created in top-left corner. When user clicks on the tray icon, container window is moved to the click location (to handle events correctly).
On KWin restart all windows are shuffled, usually KWin is able to restore ordering correctly, but for some reason not it this case. As a result black/transparent container windows are stacked above all other windows and panels.
To solve that, when container window is visible, XembedSNIProxy needs hide it again by stacking the container window below again.
BUG: 357443
FIXED-IN: 5.18.0
Test Plan:
1. Run any application with XEmbed system tray icon, do not click on the icon
2. Restart KWin
3. [Optional] Disable compositor - with disable container window is black and easier to spot
4. Expected:
a) before fix: black/transparent rectangle in the top-left corner, reacts to mouse click
b) after fix: no rectangle, mouse clicks work as expected.
Reviewers: #plasma_workspaces, #plasma, davidedmundson
Reviewed By: #plasma_workspaces, #plasma, davidedmundson
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26395
Summary:
This is not really needed here due to the size of the delegates and of
the list which is generally short. I ended up getting rid of it because
it makes this config module crash somewhere in the QML runtime when the
user clicks reset. I couldn't track down why DelegateRecycler does this
so far.
Reviewers: #plasma, mart, davidedmundson, crossi, bport
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26241
Summary:
Change the qalculate krunner module to print only significant digits instead of an interval as this is sometimes unreadable.
BUG: 412627
FIXED-IN: 5.18.0
Test Plan:
Before this change:
{F7828390}
After:
{F7828389}
Reviewers: broulik, #plasma
Reviewed By: broulik, #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26149
It is only attached to the delegate item itself.
Broke when it got shuffled around when swiping got added.
Differential Revision: https://phabricator.kde.org/D26422
Summary:
Save quite a few KFileItem instanciation.
On each document hovering several KFileItem were instanciated for nothing.
Reviewers: #frameworks, apol
Reviewed By: apol
Subscribers: apol, plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26416
Especially when you're almost out of juice you don't want to be wasting it blinking an icon causing excess repaints.
Also considering the critical percentage is hardcoded to 5% here and doesn't follow PowerDevil's setting.
Given battery critical notification is persistent and always on top of everything these days,
you can't really miss the fact that you're almost ouf of power.
Differential Revision: https://phabricator.kde.org/D26434
AppData contains a QIcon with pixmap data gathered from KWindowSystem::icon which we can just use.
Only if, for some reason, the icon is really null, try getting it from KWindowSystem on demand.
Differential Revision: https://phabricator.kde.org/D26086
Summary:
The path of places is not very informative to users
Follow the path of D26121
Test Plan: {F7828297}
Reviewers: ngraham, #plasma, #vdg, hein
Reviewed By: ngraham, #plasma, #vdg, hein
Subscribers: hein, plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26147
Summary:
Use QQC2.Slider, so that we have a moved signal. This way we can only
issue new brightnesses when the user actually interacts with it.
Don't adapt to the system brightness until we have finished interacting
with it.
Test Plan:
Manual testing, flickering is very much reduced both when scrolling over the
compact plasmoid as well as the slider.
Reviewers: #plasma, broulik
Reviewed By: #plasma, broulik
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26035
Summary:
I'm trying to implement the idea of @davidedmundson from this https://phabricator.kde.org/T10172 with my limited knowledge of C++/Qt/KDE/programming. This is the start of a new class so it could lack many things but it's usable.
- It allows clients to set the available screen rect/region from outside through dbus.
- The final available screen rect/region would be the intersection of all set values.
- When a client dies, its values would be removed automatically.
Test Plan:
+ qdbus-qt5 org.kde.plasmashell /StrutManager test org.kde.lattedock LVDS-1 0 24 1366 768
+ After killing latte-dock, the values are got only from plasmashellCorona.
{F7820969}
Reviewers: #plasma, mvourlakos, davidedmundson, mart, apol
Reviewed By: #plasma, davidedmundson
Subscribers: davidre, plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D25807
The previous QIcon::name() check wasn't sufficient as it only works with theme icon names (e.g. kate)
but not icons created from an absolute path, e.g. /some/special/icon/location/kate.png.
The latter is usually the case for containerized apps which have a proper application desktop file
installed but their icon in some path within the application image or some container daemon location.
Since we already store the information for whether we had to fall back to using the actual window pixmap,
check for this before trying to serialize icon pixmap data.
Differential Revision: https://phabricator.kde.org/D25762
This reverts commit 4000bd5db5.
Meta+space was already used for Change Keyboard Layout, which was not
known by me or the patch's reviewers. However users have started to
complain about this. Reverting to un-break people.
Sorry everyone!
Summary:
Currently, the common config window messages are neither extracted nor used in all 5 widgets. It is not possible to add several translation catalogs for a single QML widget thus the messages are extracted into all 5 widget catalogs.
Idea of the fix by Victor Ryzhykh.
Test Plan: Extraction tested to work as expected, translation should work too.
Reviewers: #localization, #plasma, davidedmundson, ngraham
Reviewed By: #plasma, davidedmundson, ngraham
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D26192
Allows to control the behavior of those applications rather than flat-out ignoring them.
Adjust default configuration to blacklist them, though.
Differential Revision: https://phabricator.kde.org/D26160
This adds a quick reply feature with a text field inline in the notification popup.
An action named "inline-reply" will spawn the text field and a NotificationReplied signal is emitted then.
There's additional kde hints for changing the placeholder text (defaults to "Type a reply..."),
submit button text (defaults to "Send") and submit button icon name (defaults to "document-send",
that paper aeroplane icon).
An application could have multiple inhibitions. Release all of them when the service unregisters (process dies).
Differential Revision: https://phabricator.kde.org/D26125
This can be useful for notifications that confirm direct user interaction, such as successful Bluetooth pairing,
Screenshot creation, or plasmoid removal.
It only overrules filtering based on urgency, disabling notifications for that application altogether will still hide them.
Such a notification also won't increase the unread counter as it can be assumed when they do something, e.g. click on "Pair",
they will notice that the pairing progress dialog went away and a confirmation notification is shown.
Differential Revision: https://phabricator.kde.org/D25935
When a plasmoid collapses in Active state, the popup is hidden.
In Passive state, however, one is returned to the hidden items list, as it cannot tell whether
the plasmoid collapsed on its own or because the user clicked the same one in the hidden items list again.
When we clear notifications, plasmoid state updates immediately as a result of model signals,
and when we then try to collapse the plasmoid we're already hidden and just returned to the hidden items list.
This patch addresses this by delaying the state change slightly.
Futhermore, don't collapse the plasmoid when already hidden as that would just return the user to the hidden items list.
Differential Revision: https://phabricator.kde.org/D26091
Nate Graham
David Edmundson
Konrad Materka
Kevin Ottens
Méven Car
Aleix Pol
Lukas Pichler
Kai Uwe Broulik
Antonio Rojas
l10n daemon script
Méven Car
Tranter Madi
Laurent Montel
Yuri Chornoivan
George Vogiatzis