Carlos Garnacho [Mon, 18 Jan 2016 12:31:35 +0000 (13:31 +0100)]
wayland: Unset button modifiers on pointer enter
Due to implicit grabs, we basically can guarantee that the pointer
won't have any buttons pressed at the time of wl_pointer.enter.
Seems like a good place to unset any button modifiers that might
have been left stale by compositor grabs.
Carlos Garnacho [Wed, 13 Jan 2016 20:00:34 +0000 (21:00 +0100)]
wayland: Implement DnD actions as per wl_data_device v3
Implement as a managed GdkDragContext, which actually fits nicely
with the drag-and-drop model in wayland.
Carlos Garnacho [Wed, 13 Jan 2016 19:57:47 +0000 (20:57 +0100)]
wayland: Add internal gdk_wayland_seat_set_global_cursor()
This can be used for cases (like DnD) where there isn't necessarily
a grab, but we want a global pointer cursor.
Carlos Garnacho [Wed, 13 Jan 2016 19:54:43 +0000 (20:54 +0100)]
gtkdnd: Reset widget controllers when starting DnD.
Otherwise there's no full guarantees that this will happen.
Carlos Garnacho [Wed, 13 Jan 2016 19:53:54 +0000 (20:53 +0100)]
gtk: Add private gtk_widget_reset_controllers()
This is a one-shot call that can be called on the places we
know we're taking input away.
Carlos Garnacho [Fri, 8 Jan 2016 20:31:51 +0000 (21:31 +0100)]
gtkdnd: Optionally use gdk_drag_context_manage_dnd()
When this is in use, there's essentially a bunch of dead code here.
When all backends are ported, we'll be able to remove grab/cursor
management plus a bunch of source-side event handlers.
Carlos Garnacho [Fri, 8 Jan 2016 20:24:52 +0000 (21:24 +0100)]
x11: Implement gdk_drag_context_manage_dnd()
This includes managing input events and source-side DND events,
as well as setting the appropriate cursor and emitting the signals
that are expected in this mode of operation.
Carlos Garnacho [Wed, 13 Jan 2016 19:51:19 +0000 (20:51 +0100)]
gdk: Add gdk_drag_get_cursor()
This function (most similar to gtk_drag_get_cursor() helps figure out
the right cursor that applies to a given action. To be used by the
various backends.
Carlos Garnacho [Fri, 8 Jan 2016 20:22:47 +0000 (21:22 +0100)]
gdk: Run DnD internal handlers before the main event handler
We'll be stealing those to GTK+, if the GdkDragContext manages
the DnD operation.
Carlos Garnacho [Fri, 8 Jan 2016 20:03:01 +0000 (21:03 +0100)]
gdk: Allow internal management of source-side DnD
We've traditionally left GTK+ to handle the input side of things,
letting GDK only manage the windowing-specific messaging. This
way of splitting responsibilities is not compatible however with
some backends, we must fold then input management at the DnD stage
into GDK (and backends) domain.
The gdk_drag_context_manage_dnd() call is meant to be the entry
point for this mode of operation, if the drag and drop operation
becomes managed, the caller (i.e. gtkdnd.c) doesn't need to perform
grabs, nor manage input events itself.
As a consequence of this, different aspects now belong to the
backend GdkDragContext implementation:
- Because the caller doesn't see keyboard events anymore,
keyboard navigation must be managed in GDK, so is the decision
of the current action based on modifiers/button pressed.
- Because the caller won't see input events in general, the lifetime
of the drag and drop operation is now communicated through the
::drop-performed, ::dnd-finished and ::cancel events
- Because the caller doesn't participate anymore on the action
being chosen, the pointer cursor must be set by the backend.
The caller is rather notified of the final action through the
::action signal.
The caller is still responsible of dealing with the corresponding
GdkSelection, ensuring its ownership and communicating the supported
mimetypes.
Benjamin Otte [Tue, 19 Jan 2016 13:11:36 +0000 (14:11 +0100)]
widget: Call gdk_window_mark_paint_from_clip() again
The proper window to call it is the event window, as the call itself
ignores non-native windows anyway.
Marek Černocký [Tue, 19 Jan 2016 11:52:58 +0000 (12:52 +0100)]
Fixed typo childen->children
Matthias Clasen [Tue, 19 Jan 2016 04:40:38 +0000 (23:40 -0500)]
gtk-demo: Add a progress bar to foreigndrawing
This is another commonly requested widget.
Benjamin Otte [Tue, 19 Jan 2016 03:33:09 +0000 (04:33 +0100)]
container: Properly reorder no-window children
... that are setup with gtk_widget_set_parent_window().
Fixes scrollbars not being drawn in GtkScrolledWindow.
Benjamin Otte [Tue, 19 Jan 2016 03:24:32 +0000 (04:24 +0100)]
widget: Add forgotten push_group code
... and remove the also forgotten void function that lingered around
with it.
Fixes opacity=0 parts like inactive spinners or sort indicators in
treeview headers being drawn since last commit.
Oops.
Benjamin Otte [Tue, 19 Jan 2016 01:42:58 +0000 (02:42 +0100)]
widget: Redo drawing code
Previously, we had a special cae to draw subwindows of widgets.
This is not necessary as conformant widgets should be able to properly
render themselves when all windows need to be painted.
From now on assume that is the case.
We therefore paint nonnative GDK windows "inline" by just returning TRUE
for gtk_cairo_should_draw_window() for those windows.
This speeds up hilighting different rows in the listbox gtk-demo example
tremendously (by a factor of 10 or more) as the previous code was
O(<number of non-window subwidgets> *
<number of subwindows>) which in the listbox example were ~15,000 and
~2,000 respectively.
Benjamin Otte [Mon, 18 Jan 2016 23:53:55 +0000 (00:53 +0100)]
actionbar: Don't forall() widgets twice
When using forall(), only list the revealer, which lists the box
containing all the children. When using foreach(), bypass revealer and
box and list all children added to the box.
Matthias Clasen [Tue, 19 Jan 2016 01:20:32 +0000 (20:20 -0500)]
Updates
Matthias Clasen [Mon, 18 Jan 2016 19:54:25 +0000 (14:54 -0500)]
text view: Expose cursor blink control
Derived classes like GtkSourceView with their own ::key-event
handler need access to this, in order to make their keynav
as nice as the builtin one, wrt to caret visibility.
https://bugzilla.gnome.org/show_bug.cgi?id=760748
Carlos Garnacho [Thu, 14 Jan 2016 19:24:28 +0000 (20:24 +0100)]
x11: Set event->scroll.is_stop
We still figure this out from 0/0 scroll events. This method is
not intended to last forever, but it's something we can cling to
so far.
https://bugzilla.gnome.org/show_bug.cgi?id=756729
Peter Hutterer [Wed, 21 Oct 2015 00:14:40 +0000 (10:14 +1000)]
wayland: add gdk_event_is_scroll_stop_event()
And use it to handle kinetic scrolling in the GtkScrolledWindow.
However, dropping the delta check causes the X11-based kinetic
scroll to break since we don't have the stop event here. Correct handling of
xf86-input-libinput-based scroll events is still being discussed.
https://bugzilla.gnome.org/show_bug.cgi?id=756729
Peter Hutterer [Tue, 20 Oct 2015 03:21:54 +0000 (13:21 +1000)]
wayland: add support for wl_pointer frame/axis_source/axis_discrete/axis_stop
This adds support for the new wl_pointer events available in v5.
The wl_pointer.axis_source events can be ignored for the purposes here, the
main reason they exist is so that the combination of axis_source=finger and
axis_stop triggers kinetic scrolling. We don't need to care about the source,
axis_stop is enough for us to tell us when we're scrolling.
The wl_pointer.frame events group events together and is intended as a
mechanism to coalesce events together. This for example allows us to now
send a single GTK scroll event for a diagonal scroll. Previously, the two
wl_pointer.axis events had to be handled separately.
The wl_pointer.axis_discrete event sends mouse wheel clicks where
appropriate, and is translated into up/down/left/right scroll events.
https://bugzilla.gnome.org/show_bug.cgi?id=756729
Matthias Clasen [Mon, 18 Jan 2016 19:31:37 +0000 (14:31 -0500)]
Drop a reference to regions
Regions don't work anymore, so don't recommend them in the docs.
Matthias Clasen [Mon, 18 Jan 2016 18:54:20 +0000 (13:54 -0500)]
im context simple: Protect shared data structures
Since
a41f02f9b1843e0f0085f801430e55f413a9bf9c, GtkIMContextSimple
uses threads to load X Compose files. It does that every time a new
im context object is initialized, so we can easily end up with multiple
threads accessing the shared global_tables list at the same time.
Use a lock to prevent that.
https://bugzilla.redhat.com/show_bug.cgi?id=
1276432
Benjamin Otte [Mon, 18 Jan 2016 17:05:53 +0000 (18:05 +0100)]
cssnode: Don't leak the style cache on destruction
Timm Bäder [Mon, 18 Jan 2016 16:22:09 +0000 (17:22 +0100)]
gdkdisplay-x11: Plug memory leak
Matthias Clasen [Mon, 18 Jan 2016 16:05:53 +0000 (11:05 -0500)]
text view: Update selection node visibility
Make the node invisible if we don't have a selection.
Matthias Clasen [Mon, 18 Jan 2016 15:52:26 +0000 (10:52 -0500)]
entry: Expand the css node docs
Mention the context menu.
Matthias Clasen [Mon, 18 Jan 2016 15:51:54 +0000 (10:51 -0500)]
text view: Expand CSS node docs
Mention the selection node, and the context menu.
Matthias Clasen [Mon, 18 Jan 2016 14:06:12 +0000 (09:06 -0500)]
Forgotten file
Benjamin Otte [Mon, 18 Jan 2016 14:01:49 +0000 (15:01 +0100)]
notebook: Don't crash when dnd'ing tabs out
https://bugzilla.gnome.org/show_bug.cgi?id=760754
Benjamin Otte [Mon, 18 Jan 2016 14:01:38 +0000 (15:01 +0100)]
accellabel: Remove unused variable
Matthias Clasen [Mon, 18 Jan 2016 13:54:35 +0000 (08:54 -0500)]
gdk: Clarify docs about scale
Explicitly state where we return application pixels as opposed
to device pixels.
Matthias Clasen [Mon, 18 Jan 2016 13:36:53 +0000 (08:36 -0500)]
Trivial typo fix
Matthias Clasen [Mon, 18 Jan 2016 12:39:56 +0000 (07:39 -0500)]
Add more CSS node tests for notebooks
Timm Bäder [Mon, 18 Jan 2016 11:46:18 +0000 (12:46 +0100)]
entry: Deprecate shadow-type
Things like that should be handled using css. Also, the property value
is unused inside GtkEntry.
Руслан Ижбулатов [Wed, 22 Apr 2015 19:21:53 +0000 (19:21 +0000)]
W32: Detect ticks wraparound
https://bugzilla.gnome.org/show_bug.cgi?id=748327
Matthias Clasen [Mon, 18 Jan 2016 05:29:29 +0000 (00:29 -0500)]
Add CSS node tests for checkbutton rtl flipping
Matthias Clasen [Mon, 18 Jan 2016 05:24:53 +0000 (00:24 -0500)]
Add css node tests for expander rtl flipping
Matthias Clasen [Mon, 18 Jan 2016 05:19:31 +0000 (00:19 -0500)]
expander: Flip the alignment of the vertical box
To ensure that the title moves to the other side as expected
in RTL, use GTK_ALIGN_START/END instead of GTK_ALIGN_FILL
when packing the title gadget into the vertical box, and
flip the alignment when the text direction changes.
Matthias Clasen [Mon, 18 Jan 2016 05:11:54 +0000 (00:11 -0500)]
expander: Handle rtl changes again
Now that the box gadget has api for it, use it here to
change the order of the gadgets for right-to-left.
Matthias Clasen [Mon, 18 Jan 2016 05:04:23 +0000 (00:04 -0500)]
checkbutton: Handle rtl changes again
Now that the box gadget has api for it, use it here to
change the order of the gadgets for right-to-left.
Matthias Clasen [Mon, 18 Jan 2016 05:03:39 +0000 (00:03 -0500)]
box gadget: Add a function to reverse children
This can be used to handle text direction changes.
Benjamin Otte [Mon, 18 Jan 2016 04:04:12 +0000 (05:04 +0100)]
reftests: Fix up for easy CSS changes
This is mostly search and replace ala
GtkButton => button
GtkWindow => window
.button => button
or removing style properties that aren't used anymore like
-GtkButton-default-border: 0
Benjamin Otte [Mon, 18 Jan 2016 01:32:19 +0000 (02:32 +0100)]
reftests: Remove test
We no longer match subtypes, so remove the test for it.
Matthias Clasen [Mon, 18 Jan 2016 02:16:33 +0000 (21:16 -0500)]
checkbutton: Use a box gadget
Now that builtin icons have a baseline, we can just use
a box gadget with baseline alignment for the indicator
and the label.
Matthias Clasen [Mon, 18 Jan 2016 02:02:18 +0000 (21:02 -0500)]
builtin icon: set a baseline
Most uses of builtin icons (check and radio buttons,
expanders, etc) are placed next to labels, so they should
be properly positioned wrt to the baseline. Lacking anything
better, give the builtin icons a baseline that places the
center of the icon at the strikethrough position.
Matthias Clasen [Mon, 18 Jan 2016 00:48:40 +0000 (19:48 -0500)]
box gadget: Deal with optional arguments
The baseline arguments to the preferred size callback can
be NULL. Don't crash if they are.
Matthias Clasen [Sun, 17 Jan 2016 20:35:14 +0000 (15:35 -0500)]
Trivial formatting fix
Matthias Clasen [Sun, 17 Jan 2016 20:26:14 +0000 (15:26 -0500)]
Document restrictions on gtk_style_context_get
This makes it official that passing a different state
is not generally supported in the getters for CSS properties.
Matthias Clasen [Sun, 17 Jan 2016 19:35:58 +0000 (14:35 -0500)]
notebook: show tabs when asked to
gtk_notebook_set_show_tabs must make the header gadget
visible when show_tabs is TRUE, not the other way around.
https://bugzilla.gnome.org/show_bug.cgi?id=760596
Matthias Clasen [Sun, 17 Jan 2016 19:08:22 +0000 (14:08 -0500)]
accel label: Fix RTL handling
The 'mad hack' that GtkAccelLabel used to affect the GtkLabel
draw function broke with the introduction of gadgets, since
the positioning is no longer relative to the widgets' allocation
at the time of the call, but rather to the gadgets allocation.
Instead of coming up with an even madder hack to keep this
working, give the GtkLabel draw function knowledge about accel
labels.
https://bugzilla.gnome.org/show_bug.cgi?id=760663
Matthias Clasen [Sun, 17 Jan 2016 18:23:49 +0000 (13:23 -0500)]
inspector: Make cursor-size combo wider
Cursor sizes typically have 2 digits, lets not clip them.
Matthias Clasen [Sun, 17 Jan 2016 17:13:59 +0000 (12:13 -0500)]
Expand window style class documentation a bit
Matthias Clasen [Sat, 16 Jan 2016 18:24:56 +0000 (13:24 -0500)]
Adwaita: Keep entry progress visible
When the entry is focused, make the progress one pixel wider
so it doesn't get swallowed by the focus outline.
Benjamin Otte [Sat, 16 Jan 2016 22:15:11 +0000 (23:15 +0100)]
stylechange: Do the right thing when old and new style are equal
We don't need to compare all their CSS values to figure out that nothing
changed. We know that.
Benjamin Otte [Sat, 16 Jan 2016 03:55:58 +0000 (04:55 +0100)]
cssnode: Redo GTK_DEBUG=no-css-cache
Previously this setting would just invalidate the whole CSS tree and
thereby hopefully avoid all cache usage.
Now, we actually don't cause extra invalidations anymore, but instead
avoid ever inserting anything into the cache when this setting is set.
Benjamin Otte [Sat, 16 Jan 2016 03:48:58 +0000 (04:48 +0100)]
cssnode: Implement the style cache
This essentially copies the previous cache implementation. With one
caveat: It is now attached to and maintained by the CssNode, not by the
CssStyle.
And this is important because styles may be reused in incompatible
situations which would cause cache collisions and lead to broken CSS in
weird situations.
Benjamin Otte [Fri, 15 Jan 2016 22:27:51 +0000 (23:27 +0100)]
cssnode: Split out the style cache
For now, the split out style cache doesn't cache anything. This is
mostly to make sure that bisections of wrong caching behavior will
bisect down to the commit that actually adds caching.
Lapo Calamandrei [Sat, 16 Jan 2016 03:16:55 +0000 (04:16 +0100)]
Adwaita: stop axing toolbar separators
Lapo Calamandrei [Sat, 16 Jan 2016 03:15:30 +0000 (04:15 +0100)]
Adwaita: style textview derived widgets
Matthias Clasen [Sat, 16 Jan 2016 01:08:57 +0000 (20:08 -0500)]
color scale: Improve themability
Add the .color class, so themes can style scale.color.
Clip the trough rendering to the content path, so that
themes can set rounded corners.
Matthias Clasen [Sat, 16 Jan 2016 00:39:12 +0000 (19:39 -0500)]
treeview: Use foreground color to draw dnd marker
This is still not perfect, but makes the dnd marker better visible
in HighContrastInverse.
Matthias Clasen [Sat, 16 Jan 2016 00:11:03 +0000 (19:11 -0500)]
frame: Don't overlap label and child
We didn't do this previously, so don't start now.
https://bugzilla.gnome.org/show_bug.cgi?id=760528
Lapo Calamandrei [Fri, 15 Jan 2016 23:55:59 +0000 (00:55 +0100)]
Adwaita: combo style cleanup and fixes
fixes https://bugzilla.gnome.org/show_bug.cgi?id=760524
Matthias Clasen [Fri, 15 Jan 2016 22:55:44 +0000 (17:55 -0500)]
combo: Add .linked class when appropriate
Add the .linked style class to combo boxes which have
an entry, since that is what they are.
Matthias Clasen [Fri, 15 Jan 2016 22:47:08 +0000 (17:47 -0500)]
combobox: Document the popup window CSS node
Matthias Clasen [Fri, 15 Jan 2016 22:36:13 +0000 (17:36 -0500)]
application: Protect against double shutdown
It is not something that can happen normally.
But if it does, we shouldn't crash.
https://bugzilla.gnome.org/show_bug.cgi?id=760680
William Hua [Fri, 15 Jan 2016 19:09:33 +0000 (14:09 -0500)]
mir: add default GdkSeat for the default GdkDisplay
Matt Watson [Tue, 12 Jan 2016 02:40:53 +0000 (18:40 -0800)]
widget: remove queue_resize in set_sensitive
Calling gtk_widget_propagate_state is enough to queue a resize if
needed
See
da7a4089fe82981d1412aaa295e76d47a81ad6df
Rui Matos [Fri, 15 Jan 2016 17:35:25 +0000 (18:35 +0100)]
wayland: Fix an endless loop
Lapo Calamandrei [Fri, 15 Jan 2016 17:33:38 +0000 (18:33 +0100)]
Adwaita: remove messagedialog label squashing hack
since Matthias fixed this properly.
Matthias Clasen [Fri, 15 Jan 2016 16:58:51 +0000 (11:58 -0500)]
message dialog: Prevent empty title from being shown
We don't want to grow the titlebar by showing an empty label.
Lapo Calamandrei [Fri, 15 Jan 2016 16:56:49 +0000 (17:56 +0100)]
Adwaita: some dialog love
addresses https://bugzilla.gnome.org/show_bug.cgi?id=760525
Matthias Clasen [Fri, 15 Jan 2016 15:29:20 +0000 (10:29 -0500)]
expander: Destroy gadgets after child widgets
With child widgets now being sorted into box gadgets, bad things
happen if those gadgets disappear prematurely.
Benjamin Otte [Fri, 15 Jan 2016 11:58:14 +0000 (12:58 +0100)]
notebook: Destroy gadgets after chaining up
While chaining up, we might still call into the notebook code. And that
code expects the gadgets to be there.
https://bugzilla.gnome.org/show_bug.cgi?id=760640
Matthias Clasen [Fri, 15 Jan 2016 14:41:44 +0000 (09:41 -0500)]
expander: Convert to use box gadgets
Use a vertical box gadget for the overall expander, and a
horizontal one for the title row. This lets us get rid of
all the custom allocation code here.
Matthias Clasen [Fri, 15 Jan 2016 13:35:39 +0000 (08:35 -0500)]
expander: Deprecate spacing property
The same can be achieved with margins on the child, and
the spacing property gets in the way of converting the
expander to use box gadgets.
Matthias Clasen [Fri, 15 Jan 2016 11:47:31 +0000 (06:47 -0500)]
box gadget: Implement cross-axis alignment
So far, the box gadget is always allocating all children the
full size in the cross axis. This behavior corresponds to the
align-items: stretch behavior in
https://www.w3.org/TR/css-flexbox-1/#align-items-property
This commit implements the other modes described there.
While widgets have halign/valign properties that we can use for
this, the API for inserting gadgets has to change to take an
extra align parameter. All callers have been updated to pass
GTK_ALIGN_FILL, since that corresponds to the previous behavior.
https://bugzilla.gnome.org/show_bug.cgi?id=760668
Piotr Drąg [Fri, 15 Jan 2016 14:18:32 +0000 (15:18 +0100)]
Updated POTFILES.skip
Ignacio Casal Quinteiro [Fri, 15 Jan 2016 19:45:45 +0000 (11:45 -0800)]
win32: get rid of _gdk_screen global variable
The screen is part of the display so it should be belong to it.
Ignacio Casal Quinteiro [Fri, 15 Jan 2016 19:53:51 +0000 (11:53 -0800)]
win32: fix warning
Tom Schoonjans [Sun, 10 Jan 2016 04:07:07 +0000 (09:37 +0530)]
Windows library manager fix
The Makefile.am's had a dependency on incorrectly named libtool
archives for the generation of the .lib files.
https://bugzilla.gnome.org/show_bug.cgi?id=760388
Alberts Muktupāvels [Tue, 12 Jan 2016 19:34:56 +0000 (21:34 +0200)]
gtkwidgetpath: update gtk_widget_path_to_string
From gtk_widget_path_iter_set_object_name documentation:
"When set, the object name overrides the object type when matching CSS"
Update gtk_widget_path_to_string to match this behaviour.
Eric Williams [Wed, 13 Jan 2016 13:29:34 +0000 (08:29 -0500)]
Fix typo in previous documentation fix for gtk_css_provider_load_from_data()
Matthias Clasen [Thu, 14 Jan 2016 21:06:16 +0000 (16:06 -0500)]
Document -gtk-recolor syntax
Matthias Clasen [Thu, 14 Jan 2016 18:48:43 +0000 (13:48 -0500)]
Add a way to load recolored icons in CSS
This can come in handy to load images that are not in an icon theme,
but still benefit from symbolic recoloring.
https://bugzilla.gnome.org/show_bug.cgi?id=760536
Matthias Clasen [Thu, 14 Jan 2016 18:49:08 +0000 (13:49 -0500)]
icon theme: Export recoloring function
This will be used in subsequent commits.
https://bugzilla.gnome.org/show_bug.cgi?id=760536
Matthias Clasen [Thu, 14 Jan 2016 17:53:50 +0000 (12:53 -0500)]
icon theme: Add a way to create GtkIconInfos from files
This is sometimes needed, and calling into actual icon theme
code just for it is confusing - the resulting icon does not
depend on the icon theme at all.
https://bugzilla.gnome.org/show_bug.cgi?id=760536
Lapo Calamandrei [Thu, 14 Jan 2016 16:51:49 +0000 (17:51 +0100)]
Adwaita: pathbar style cleanup
Florian Müllner [Thu, 14 Jan 2016 13:45:28 +0000 (14:45 +0100)]
cssstyleproperty: Expose min-width/height
Lapo Calamandrei [Thu, 14 Jan 2016 15:26:22 +0000 (16:26 +0100)]
Adwaita: remove superfluous selector on keycap styling
now that keys are made directly with labels istead of frames.
Timm Bäder [Wed, 13 Jan 2016 18:24:46 +0000 (19:24 +0100)]
boxgadget: Don't use g_assert
Use g_return_if_fail instead so nothing abort()s.
Timm Bäder [Wed, 13 Jan 2016 18:14:33 +0000 (19:14 +0100)]
entry: Don't draw icons twice
Timm Bäder [Wed, 13 Jan 2016 18:11:24 +0000 (19:11 +0100)]
entry: Remove unused code
Timm Bäder [Wed, 13 Jan 2016 18:07:20 +0000 (19:07 +0100)]
label: Remove some unneeded conditionals
We are never passing NULL here.
Timm Bäder [Sat, 9 Jan 2016 16:15:29 +0000 (17:15 +0100)]
shortcutlabel: Use a label directly
We can everything we need without the frame around it now.
Руслан Ижбулатов [Sun, 27 Dec 2015 05:05:39 +0000 (05:05 +0000)]
GDK W32: Don't use WM-drawn shadow for tooltips
This prevents WM from drawing shadows around tooltip windows,
which, in Adwaita, should have no shadow and are CSD-ish (which means
that tooltip window is larger than it looks, and WM draws the shadow
only on the outside, leaving a gap between the visible tooltip edge and
the shadow).
https://bugzilla.gnome.org/show_bug.cgi?id=759898
Paolo Borelli [Thu, 17 Dec 2015 23:36:57 +0000 (00:36 +0100)]
win32: remove _gdk_root global variable
Руслан Ижбулатов [Thu, 14 Jan 2016 12:51:26 +0000 (12:51 +0000)]
Ensure that the fstat fix is not applied on x86_64