GitLab

Specifically it is avoided to be toggled if:
- Just received focus (in order to preserve OSK state across focus changes)
- Moving cursor around. Still allow some jitter as perfect accuracy is not
  possible.

Closes: https://gitlab.gnome.org/GNOME/gtk/issues/1277

Fix -Wincompatible-pointer-types warnings

See merge request GNOME/gtk!523

Closes: https://gitlab.gnome.org/GNOME/gtk/issues/1600

SpinButton: Fix wrong background/icon colours in button:backdrop:disabled

See merge request GNOME/gtk!489

filechoosernative: Implement gtk_file_chooser_set_filter()

See merge request GNOME/gtk!435

Fix long combo box positioning

See merge request GNOME/gtk!514

Let people know that up to Visual Studio 2017 is supported, and on
enabling additional features which are not enabled by default.

Make things consistent across the board.  Only the .sln files and the
various README files under build/win32 should have Windows CR/LF line
endings.

This file is generated during 'make dist', so we should not need to
track it in GIT.

Adapt the Visual Studio project files to output the introspection files
in the same directories where the built binaries are located from the
previous patch.

Also, make the gtk3-introspect project dependent on the gdk-3 and gtk-3
projects only, so that we can build the introspection files without
needing to finish the whole build process.  In order to "install" the
built introspection files, the gtk3-install project is now where this is
being done.  Note that the introspection builds is still not built by
default at this point.

To avoid confusion, have the NMake Makefiles output the built introspection
files in the same location where the binaries are built for the project
files, according to the Visual Studio version, platform and configuration
where the build is carried out.

Also make generating the introspection NMake snippet portion more robust to
source additions and removals by checking on Makefile changes too.

When a popup is placed using move_to_rect(), it'll get feedback about
the position and size it got assigned. We use this feedback to update
the scroll offset, but while doing so, if the visibility of the arrow
changed, we didn't adapt the offset accordingly.

Fix this by offsetting the provided offset by the height of the arrow,
if it was made visible as a side effect of the scroll offset change
triggered by the feedback.

Related: mutter#105
Closes: #1463

A menu will be clamped to the work area as a side effect of the
move_to_rect() logic if the resize anchor flags was set. For it to work
a second time, the initial size needs to be the actual menu size before
being clamped again. Achieve this by forcing a size recalculation before
showing the menu.

Don't constrain the initial menu size by the work area of some monitor;
instead let the move_to_rect() logic in the backend do the constraining.
This fixes two things:

1) The anchor delta provided to the backend will not be invalid. The
delta is calculated by looking at the active menu item, calculating the
offset given that, but since we clamped the window size before showing
the window, the delta became invalid. This caused visible issues when
the delta was large enough to make the initially calculated popup window
geometry to be placed outside the geometry of the parent window, which
is a violation of the Wayland protocol.

2) The scroll offset to be correct when receiving the positioning
feedback. While the scroll offset was based on the pre-clamped window
size, the feedback, which was used to calculate the new offset, was not,
causing the scroll offset to be clamped as well.

If the size was constrained by the xdg_positioner mechanisms, we handle
the resize by resizing the popup window. What we shouldn't do is
hide/show the popup window so avoid that.

We need to tell the portal what filter is supposed to be selected by
default, or it will just pick the first one, which could be wrong and
annoying.

This will require updated xdg-desktop-portal and xdg-desktop-portal-gtk
to work properly.

Fixes #1492

about dialog: Use ::activate-link for license

See merge request GNOME/gtk!510

This should behave the same as all the other links
in this dialog.

Closes: #1561

Add fallbacks for GtkSwitch state labels

See merge request GNOME/gtk!508

While the IEC power symbols have been part of Unicode since version 9.0,
released in 2016, not every font supports them.

We can use the old symbols as a fallback, as they seem to have the
better coverage, if not the best appearance.

We are casting from GtkFileChooserNative to... GtkFileChooserNative.
Boo.

gtk_file_chooser_set_filter() doesn't work for GtkFileChooserNative. The
code forwards added and removed filters to the delegate dialog, but
doesn't do anything to set the selected one, so the wrong one gets
chosen. So fix that.

This only fixes the fallback dialog. The portal will be fixed in a
subsequent commit.

Partial fix for #1492

Cell rendererers don't support <style> in ui files.

Instead of from the IMContextQuartz's client window because the former
is the event window where the text will be inserted. In some cases
they're different and the text may be discarded (because the client
window isn't editable) or misplaced.

Fixes Bug 707945.

Use power symbols for GtkSwitch labels

See merge request GNOME/gtk!503

Both the "on" and "off" IEC states have a corresponding Unicode glyph,
starting from Unicode 9.0.

listbox: set selected_row before emitting signal

See merge request GNOME/gtk!479

switch: Always use I/O

See merge request GNOME/gtk!501

This allows for smaller switches, and goes better with
modern themes that have round switches.

Revert "gdk: deactivate/activate surface on keyboard grabs"

See merge request GNOME/gtk!484