This may still not raise the application correctly on some WMs, because
they don't allow disturbing the user with windows popping up. I don't
think we want to work around that.
fixes#84fixes#113
removed from:
- emoji panel scrollbars
- emoji category labels
- splitter image handles
- textfield setTextColor impl
small change to the category separator label for better contrast /
readability in dark mode.
removed setTextColor completely from TextField. Doesn't appear to be in
use anywhere, and focus going more toward qss themeing from qproperty
setting.
setStyleSheet was overriding default tooltip theme from the system,
causing themes like Breeze to render white text on a grey bg for
tooltips. Tooltips now render default for the system theme everywhere.
set a theme matching color on the dark theme comboboxes.
fixes#118
- Switch to {} widget initialization
- set fixed size on the Toggles so they don't scale down vertically
- add another setting section for INTERFACE
- add a placeholder value for deviceFingerprintValue_ so it can calculate the right size.
- fix warnings about `QLayout: Attempting to add QLayout "" to UserSettingsPage "",` in boxWrap
- hide widgets properly in Q_OS_MAC
- remove mainLayout_ completely
- clean up resizeEvent
This allows form fields to wrap dynamically when they are too long for
the view. This should help mobile and small screen users to configure
the app without having to force it into portrait mode, or not be able to
reach the settings fields at all.
QFileDialog's dir arg (which was set to the incoming file name from the
Matrix download) can take a full path to suggest. By prepending
QStandardPaths::DownloadLocation, it opens to the system's download
folder and proposes the filename as the download name.
Using QStandardPaths should make this work on other platforms, and from
what I read, its possible for this to return an empty string on
platforms where it doesn't support it, so this should essentially
revert to the previous functionality if Qt can't determine the system's
download location.
it was causing the full screen image overlay to appear on the "primary
display" vs. the actual display nheko is running on.
removing the move() call makes the overlay follow the window.
