gtk4.git
9 years agoentry: handle NULL cursor when updating cursors
Christian Hergert [Tue, 27 Sep 2016 21:12:43 +0000 (14:12 -0700)]
entry: handle NULL cursor when updating cursors

If we got a NULL cursor from gdk_cursor_new_from_name(), we need
to handle that when unreffing.

9 years agoentry: handle NULL cursor when releasing reference
Christian Hergert [Tue, 27 Sep 2016 21:09:02 +0000 (14:09 -0700)]
entry: handle NULL cursor when releasing reference

If we happen to have gotten NULL back for a cursor, then we
need to check before unreffing it.

9 years agotreeviewcolumn: use g_clear_object when freeing cursor
Christian Hergert [Tue, 27 Sep 2016 19:41:55 +0000 (12:41 -0700)]
treeviewcolumn: use g_clear_object when freeing cursor

There is the chance that the cursor returned NULL on systems
with missing cursors. This defensively handles that by checking
for a NULL cursor.

9 years agoMake gtk_init_with_args behave as documented
Matthias Clasen [Tue, 27 Sep 2016 19:12:47 +0000 (15:12 -0400)]
Make gtk_init_with_args behave as documented

This docs say it should return FALSE if the display can't be opened,
but it was failing.

https://bugzilla.gnome.org/show_bug.cgi?id=771959

9 years agopaned: quietly handle missing cursors
Matthias Clasen [Tue, 27 Sep 2016 18:55:13 +0000 (14:55 -0400)]
paned: quietly handle missing cursors

The row-/col-resize cursors should really be present, but if they
aren't we don't need to crash.

https://bugzilla.gnome.org/show_bug.cgi?id=771963

9 years agoFix wrong URL link in gtkshortcutswindow documentation
Gautier Pelloux-Prayer [Tue, 27 Sep 2016 13:51:50 +0000 (15:51 +0200)]
Fix wrong URL link in gtkshortcutswindow documentation

https://bugzilla.gnome.org/show_bug.cgi?id=772057

9 years ago_gtk_icon_helper_draw: get style earlier
Daniel Drake [Mon, 26 Sep 2016 19:18:59 +0000 (13:18 -0600)]
_gtk_icon_helper_draw: get style earlier

After checking for rendered_surface, the call to gtk_css_node_get_style
can invalidate the style and result in rendered_surface being set to
NULL. This was result in some icon views appearing blank on
Endless OS on armv7hl, and this error:

Gtk-CRITICAL **: gtk_css_style_render_icon_surface: assertion 'surface != NULL' failed

Call gtk_css_node_get_style earlier to ensure we always pass a valid
surface to gtk_css_style_render_icon_surface.

https://bugzilla.gnome.org/show_bug.cgi?id=765649
https://phabricator.endlessm.com/T13524

9 years agoDo not filter out small scroll event fractions
Andreas Pokorny [Sat, 13 Aug 2016 23:29:35 +0000 (01:29 +0200)]
Do not filter out small scroll event fractions

The scroll motion values are subject of batching and scaling. Either
through scaling or by using a touchpad smooth scroll motion changes
below 0.5 are possible.

https://bugzilla.gnome.org/show_bug.cgi?id=769554

Signed-off-by: Andreas Pokorny <andreas.pokorny@canonical.com>
9 years agoUpdated French translation
Charles Monzat [Mon, 26 Sep 2016 08:45:58 +0000 (08:45 +0000)]
Updated French translation

9 years agobloatpad: replace single quotes with double quotes in menus.ui
Piotr Drąg [Sat, 24 Sep 2016 16:17:34 +0000 (18:17 +0200)]
bloatpad: replace single quotes with double quotes in menus.ui

Double quotes are more friendly to some l10n tools, and this sets
a good example for application authors to follow.

9 years agoAdwaita: always use images for button backgrounds
Lapo Calamandrei [Sat, 24 Sep 2016 14:08:39 +0000 (16:08 +0200)]
Adwaita: always use images for button backgrounds

transitions gets messed up when some background-image and
background-color are in the mix.
Fixes: https://bugzilla.gnome.org/show_bug.cgi?id=769500
9 years agoUpdated Croatian translation
gogo [Sat, 24 Sep 2016 10:12:21 +0000 (10:12 +0000)]
Updated Croatian translation

9 years agoUpdated Croatian translation
gogo [Sat, 24 Sep 2016 10:07:29 +0000 (10:07 +0000)]
Updated Croatian translation

9 years agogtktreeviewcolumn: Fix typo in typedef documentation
Philip Withnall [Sat, 24 Sep 2016 04:10:13 +0000 (21:10 -0700)]
gtktreeviewcolumn: Fix typo in typedef documentation

9 years agoUpdated Hungarian translation
Balázs Meskó [Fri, 23 Sep 2016 08:50:09 +0000 (08:50 +0000)]
Updated Hungarian translation

9 years agoUpdated Brazilian Portuguese translation
Rafael Fontenelle [Fri, 23 Sep 2016 00:13:40 +0000 (00:13 +0000)]
Updated Brazilian Portuguese translation

9 years agoUpdated Swedish translation
Anders Jonsson [Thu, 22 Sep 2016 20:09:46 +0000 (20:09 +0000)]
Updated Swedish translation

9 years agoUpdated Hebrew translation
Yosef Or Boczko [Thu, 22 Sep 2016 20:06:57 +0000 (23:06 +0300)]
Updated Hebrew translation

9 years agoUpdated Slovak translation
Dušan Kazik [Thu, 22 Sep 2016 20:00:54 +0000 (20:00 +0000)]
Updated Slovak translation

9 years agoUpdated German translation
Mario Blättermann [Thu, 22 Sep 2016 18:14:38 +0000 (18:14 +0000)]
Updated German translation

9 years agoFix another type in the css docs
Matthias Clasen [Thu, 22 Sep 2016 18:01:45 +0000 (11:01 -0700)]
Fix another type in the css docs

9 years agoFix a typo in the css docs
Matthias Clasen [Thu, 22 Sep 2016 17:58:34 +0000 (10:58 -0700)]
Fix a typo in the css docs

9 years agoUpdated Kazakh translation
Baurzhan Muftakhidinov [Thu, 22 Sep 2016 16:25:29 +0000 (16:25 +0000)]
Updated Kazakh translation

9 years agoUpdated Polish translation
Piotr Drąg [Thu, 22 Sep 2016 16:17:20 +0000 (18:17 +0200)]
Updated Polish translation

9 years agoUpdated Korean translation
Changwoo Ryu [Thu, 22 Sep 2016 15:27:05 +0000 (15:27 +0000)]
Updated Korean translation

9 years agoUpdated Czech translation
Marek Černocký [Thu, 22 Sep 2016 15:26:05 +0000 (17:26 +0200)]
Updated Czech translation

9 years agoUpdated Korean translation
Changwoo Ryu [Thu, 22 Sep 2016 15:25:20 +0000 (15:25 +0000)]
Updated Korean translation

9 years agogtkplacesview: mark string as translatable
Carlos Soriano [Thu, 22 Sep 2016 14:56:51 +0000 (16:56 +0200)]
gtkplacesview: mark string as translatable

We missed it.

https://bugzilla.gnome.org/show_bug.cgi?id=771666

9 years agoAdd some missing (nullable) annotations
Rico Tzschichholz [Thu, 22 Sep 2016 11:56:47 +0000 (13:56 +0200)]
Add some missing (nullable) annotations

https://bugzilla.gnome.org/show_bug.cgi?id=771826

9 years agowidget: Don't print missing allocation warning by default
Timm Bäder [Thu, 22 Sep 2016 06:58:14 +0000 (08:58 +0200)]
widget: Don't print missing allocation warning by default

This was meant to be silenced unless expicitly requested but
G_ENABLE_DEBUG is defined by default unless --disable-debug is passed to
configure, so use G_ENABLE_CONSISTENCY_CHECKS instead which is only
defined if --enable-debug is explicitly passed.

9 years agoencodesymbolic: handle circle objects
Lapo Calamandrei [Wed, 21 Sep 2016 15:46:19 +0000 (17:46 +0200)]
encodesymbolic: handle circle objects

add circle objects to the injected style for recoloring.
Should avoid randomly colored symbolic icon bits when circles are
in the mix as in network-vpn-acquiring-symbolic for example.

9 years agoUpdate Arabic translation
Khaled Hosny [Wed, 21 Sep 2016 07:05:35 +0000 (09:05 +0200)]
Update Arabic translation

9 years agodocs: Add a diagram for GtkListBox nodes
Emmanuele Bassi [Tue, 20 Sep 2016 22:32:00 +0000 (23:32 +0100)]
docs: Add a diagram for GtkListBox nodes

GtkListBox's API reference for CSS nodes should follow the template of
other widget classes.

9 years agoUpdate Basque language
Inaki Larranaga Murgoitio [Mon, 19 Sep 2016 11:22:33 +0000 (13:22 +0200)]
Update Basque language

9 years agoUpdate Basque language
Inaki Larranaga Murgoitio [Mon, 19 Sep 2016 11:21:55 +0000 (13:21 +0200)]
Update Basque language

9 years ago3.22.0
Matthias Clasen [Mon, 19 Sep 2016 17:07:29 +0000 (10:07 -0700)]
3.22.0

9 years agoUpdated Ukrainian translation
Daniel Korostil [Mon, 19 Sep 2016 21:05:57 +0000 (00:05 +0300)]
Updated Ukrainian translation

9 years agoUpdated Danish translation
Ask Hjorth Larsen [Mon, 19 Sep 2016 20:31:14 +0000 (22:31 +0200)]
Updated Danish translation

9 years agoAdwaita: sync svg assets with css
Jakub Steiner [Mon, 19 Sep 2016 20:21:23 +0000 (22:21 +0200)]
Adwaita: sync svg assets with css

- scale slider strokes were lighter on the SVG variants

9 years agogtkwindow: Update shadow size on state change
Olivier Fourdan [Mon, 19 Sep 2016 14:11:54 +0000 (16:11 +0200)]
gtkwindow: Update shadow size on state change

Otherwise, with CSD, we could have a discrepancy where gtk uses the
right values for the shadows whereas the gdk backend still uses the old
values, leading in some cases to invalid or negative min size being
computed (which, in Wayland, leads a protocol error).

https://bugzilla.gnome.org/show_bug.cgi?id=771561

9 years agoGDK W32: Ignore autorepeated key presses on modifier keys
Jeremy Tan [Sat, 17 Sep 2016 12:46:30 +0000 (20:46 +0800)]
GDK W32: Ignore autorepeated key presses on modifier keys

The X11 backend does not send autorepeated messages for modifier keys,
and doing so prevents motion compression from working.

https://bugzilla.gnome.org/show_bug.cgi?id=771568

9 years agoGDK W32: Always process all available messages
Jeremy Tan [Sat, 17 Sep 2016 09:19:59 +0000 (17:19 +0800)]
GDK W32: Always process all available messages

The GLib main loop blocks on MsgWaitForMultipleObjectsEx to
determine if there are any incoming messages while also allowing
for background tasks to run. If all available messages are not
processed after MsgWaitForMultipleObjectsEx has signaled that
there are available, CPU usage will skyrocket.

From my limited understanding (by inspection of profiling
under Visual Studio):
Key is pressed - MsgWaitForMultipleObjectsEx unblocks, and
sends message to GDK's event handler. Some event is now queued.

g_poll unblocks, calls the g_event_dispatch which finally
resolves to gdk_event_dispatch. This then calls
_gdk_win32_display_queue_events, but since a message is already
queued, it fails to call PeekMessage and returns immediately.

At the next iteration, g_poll again calls MsgWaitForMultipleObjectsEx
which queues yet another event and returns almost immediately, since
there are events available which haven't been processed by PeekMessage.

The dispatch function is then called and the process repeats.

https://bugzilla.gnome.org/show_bug.cgi?id=771568

9 years agoUpdate British English translation
David King [Sun, 18 Sep 2016 12:12:06 +0000 (14:12 +0200)]
Update British English translation

9 years agoUpdated Kazakh translation
Baurzhan Muftakhidinov [Sun, 18 Sep 2016 08:47:39 +0000 (08:47 +0000)]
Updated Kazakh translation

9 years agoGDK W32: Set is_modifier field in GdkKeyEvent
Руслан Ижбулатов [Sat, 17 Sep 2016 15:13:45 +0000 (15:13 +0000)]
GDK W32: Set is_modifier field in GdkKeyEvent

This hardcodes Control, Alt and Shift as modifier keys.

https://bugzilla.gnome.org/show_bug.cgi?id=602773

9 years agodocs: Update gtk_window_get_size()
Emmanuele Bassi [Sat, 17 Sep 2016 11:47:39 +0000 (12:47 +0100)]
docs: Update gtk_window_get_size()

The main corpus of the documentation for gtk_window_get_size() is still
full of X11-isms, so we should port it to something that is more
backend-agnostic. Additionally, having some examples would be nice for
application authors looking at a way to appropriately use this function.

9 years agoUpdated Greek translation
Tom Tryfonidis [Fri, 16 Sep 2016 13:43:18 +0000 (13:43 +0000)]
Updated Greek translation

9 years agoUpdated Galician translations
Fran Dieguez [Fri, 16 Sep 2016 08:34:20 +0000 (10:34 +0200)]
Updated Galician translations

9 years agoUpdated Swedish translation
Anders Jonsson [Thu, 15 Sep 2016 21:54:18 +0000 (21:54 +0000)]
Updated Swedish translation

9 years agoUpdated French translation
Claude Paroz [Thu, 15 Sep 2016 12:37:43 +0000 (12:37 +0000)]
Updated French translation

9 years agoEnsure drawing context is set
Emmanuele Bassi [Thu, 15 Sep 2016 09:05:05 +0000 (10:05 +0100)]
Ensure drawing context is set

If somebody decides to use gtk_widget_set_double_buffered() in the
middle of a draw() then there's the risk of calling end_draw_frame()
with an invalid pointer.

Some overeager compilers may warn about the double_buffered bit field
changing values and leading to a potentially uninitialized variable.

In order to avoid compiler warnings or crashes, we can simply store the
value of the double_buffered bit field at the beginning of the rendering
and use that instead of the actual bit field.

https://bugzilla.gnome.org/show_bug.cgi?id=771463

9 years agoUpdated Slovak translation
Dušan Kazik [Thu, 15 Sep 2016 09:15:48 +0000 (09:15 +0000)]
Updated Slovak translation

9 years agowidget: Only warn about missing allocation if G_ENABLE_DEBUG
Timm Bäder [Wed, 14 Sep 2016 22:06:54 +0000 (18:06 -0400)]
widget: Only warn about missing allocation if G_ENABLE_DEBUG

Not all occurrences of this warning can be fixed today, so put it behind
a G_ENABLE_DEBUG flag since it still shows legitimate problems even if
some of them are false positives.

9 years agoUpdated Spanish translation
Daniel Mustieles [Wed, 14 Sep 2016 18:43:32 +0000 (20:43 +0200)]
Updated Spanish translation

9 years agoUpdated Spanish translation
Daniel Mustieles [Wed, 14 Sep 2016 18:43:22 +0000 (20:43 +0200)]
Updated Spanish translation

9 years agoImplement gdk_screen_get_monitor_scale_factor generically
Matthias Clasen [Tue, 13 Sep 2016 14:45:06 +0000 (10:45 -0400)]
Implement gdk_screen_get_monitor_scale_factor generically

This was forgotten when the other screen monitor apis were
ported to GdkMonitor.

https://bugzilla.gnome.org/show_bug.cgi?id=771349

9 years agoUpdated Brazilian Portuguese translation
Rafael Fontenelle [Wed, 14 Sep 2016 05:13:58 +0000 (05:13 +0000)]
Updated Brazilian Portuguese translation

9 years agowayland: Move and resize popup after it was configured
Jonas Ådahl [Mon, 12 Sep 2016 10:46:44 +0000 (18:46 +0800)]
wayland: Move and resize popup after it was configured

A popup may have moved and resized when configured. Make sure every
layer knows about this and call gdk_window_move_resize() with the
configured dimension and position. This won't actually move the
window, but might resize it.

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agowayland: Transform moved_to_rect result properly
Jonas Ådahl [Mon, 12 Sep 2016 10:41:36 +0000 (18:41 +0800)]
wayland: Transform moved_to_rect result properly

The result of move_to_rect, received from the xdg_popup.configure
event, needs to be translated to the correct coordinate space; that is
from real parent window geometry to coordinates relative to the gdk
window set as transient-for.

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agowayland: Use helper to translate to real parent window geometry
Jonas Ådahl [Mon, 12 Sep 2016 10:39:06 +0000 (18:39 +0800)]
wayland: Use helper to translate to real parent window geometry

Use a helper to translate a coordinate from non-real GdkWindow parent
to window geometry coordinate space of the real GdkWindow parent,
meaning the coordinate space of the GdkWindow of the parent used as a
xdg_popup parent where (0, 0) is inside of the shadow margin.

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agowayland: Don't pass parent when creating dynamic positioner
Jonas Ådahl [Mon, 12 Sep 2016 10:24:50 +0000 (18:24 +0800)]
wayland: Don't pass parent when creating dynamic positioner

When using the dynamic positioner (i.e. positioning from move_to_rect)
we can always rely on having a proper transient-for to position
relative to, so lets drop the ignored parameter.

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agowayland: Don't pass transient-for when getting real parent
Jonas Ådahl [Mon, 12 Sep 2016 10:24:14 +0000 (18:24 +0800)]
wayland: Don't pass transient-for when getting real parent

It's always derived from transient-for so no need to pass it.

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agowayland: Fix south-west anchor rect calculation
Jonas Ådahl [Mon, 12 Sep 2016 07:01:24 +0000 (15:01 +0800)]
wayland: Fix south-west anchor rect calculation

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agowayland: Don't pass non-changing state when calculating popup rects
Jonas Ådahl [Mon, 12 Sep 2016 07:00:21 +0000 (15:00 +0800)]
wayland: Don't pass non-changing state when calculating popup rects

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agowayland: Move move_to_rect related code closer together
Jonas Ådahl [Mon, 12 Sep 2016 06:53:51 +0000 (14:53 +0800)]
wayland: Move move_to_rect related code closer together

Move the code used for calculating the result of move_to_rect
(final_rect, flipped_rect etc) closer to the other move_to_rect
functions (i.e. next to create_dynamic_positioner), and let the
xdg_popup configure handler just call the calculation function.

https://bugzilla.gnome.org/show_bug.cgi?id=771117

9 years agoUpdate Catalan translation
Jordi Mas [Tue, 13 Sep 2016 17:52:46 +0000 (19:52 +0200)]
Update Catalan translation

9 years ago[l10n] update Persian translations
Arash Mousavi [Tue, 13 Sep 2016 13:38:46 +0000 (18:08 +0430)]
[l10n] update Persian translations

9 years agoUpdated French translation
Charles Monzat [Tue, 13 Sep 2016 10:18:34 +0000 (10:18 +0000)]
Updated French translation

9 years ago3.21.6
Matthias Clasen [Tue, 13 Sep 2016 04:25:15 +0000 (00:25 -0400)]
3.21.6

9 years agoUpdated Galician translations
Fran Dieguez [Tue, 13 Sep 2016 09:09:41 +0000 (11:09 +0200)]
Updated Galician translations

9 years agoUpdated Czech translation
Marek Černocký [Tue, 13 Sep 2016 08:59:29 +0000 (10:59 +0200)]
Updated Czech translation

9 years agoUpdated French translation
Erwan Georget [Mon, 12 Sep 2016 21:30:41 +0000 (21:30 +0000)]
Updated French translation

9 years agoUpdated Lithuanian translation
Aurimas Černius [Mon, 12 Sep 2016 17:49:17 +0000 (20:49 +0300)]
Updated Lithuanian translation

9 years agoAdd more options to XGETTEXT_OPTIONS in Makevars
Piotr Drąg [Mon, 12 Sep 2016 17:46:09 +0000 (19:46 +0200)]
Add more options to XGETTEXT_OPTIONS in Makevars

9 years agoUpdated Finnish translation
Jiri Grönroos [Mon, 12 Sep 2016 17:27:09 +0000 (17:27 +0000)]
Updated Finnish translation

9 years agoUpdated Hungarian translation
Balázs Meskó [Mon, 12 Sep 2016 09:06:21 +0000 (09:06 +0000)]
Updated Hungarian translation

9 years agowayland: unmap popup along with its toplevel
Olivier Fourdan [Mon, 5 Sep 2016 15:53:38 +0000 (17:53 +0200)]
wayland: unmap popup along with its toplevel

If an application umaps the toplevel from its popup callback, this can
lead to a protocol error.

Make sure we mark popup parent and use that to check if their parent is
the toplevel being unmapped in which case we shall unmap the popup first
to avoid the protocol error.

Bugzilla: https://bugzilla.gnome.org/show_bug.cgi?id=770906

9 years agoUpdated Danish translation
Ask Hjorth Larsen [Mon, 12 Sep 2016 01:16:16 +0000 (03:16 +0200)]
Updated Danish translation

9 years agoUpdated Danish translation
Ask Hjorth Larsen [Mon, 12 Sep 2016 01:13:53 +0000 (03:13 +0200)]
Updated Danish translation

9 years agoUpdated Danish translation
Ask Hjorth Larsen [Mon, 12 Sep 2016 01:11:36 +0000 (03:11 +0200)]
Updated Danish translation

9 years agoUpdated Swedish translation
Anders Jonsson [Sun, 11 Sep 2016 19:44:01 +0000 (19:44 +0000)]
Updated Swedish translation

9 years agoUpdated Russian translation
Stas Solovey [Sun, 11 Sep 2016 18:20:55 +0000 (18:20 +0000)]
Updated Russian translation

9 years agoUpdated German translation
Mario Blättermann [Sun, 11 Sep 2016 17:00:23 +0000 (17:00 +0000)]
Updated German translation

9 years agoUpdated Polish translation
Piotr Drąg [Sun, 11 Sep 2016 16:55:25 +0000 (18:55 +0200)]
Updated Polish translation

9 years agoRedo focus handling in treeview once more
Matthias Clasen [Sun, 11 Sep 2016 15:47:55 +0000 (11:47 -0400)]
Redo focus handling in treeview once more

The fix for bug 767468 had some unintended side-effects. This is
an attempt at doing the same fix (don't grab focus when we are
grab-shadowed), while avoiding the breakage, by using GTK+'s
internal tracking for grab-shadowed-ness.

https://bugzilla.gnome.org/show_bug.cgi?id=770508

9 years agoAdd a --version option to gtk3-widget-factory
Matthias Clasen [Sun, 11 Sep 2016 15:21:19 +0000 (11:21 -0400)]
Add a --version option to gtk3-widget-factory

This was missing, for no good reason.

9 years agoAdd a --version option to gtk3-demo
Matthias Clasen [Sun, 11 Sep 2016 15:04:40 +0000 (11:04 -0400)]
Add a --version option to gtk3-demo

This was missing, for no good reason.

9 years agogtk-launch: Add help string for --version
Matthias Clasen [Sun, 11 Sep 2016 14:51:19 +0000 (10:51 -0400)]
gtk-launch: Add help string for --version

This was missing for no good reason.

9 years agoUpdated Hebrew translation
Yosef Or Boczko [Sat, 10 Sep 2016 20:43:57 +0000 (23:43 +0300)]
Updated Hebrew translation

9 years agoUpdated Slovak translation
Dušan Kazik [Sat, 10 Sep 2016 14:33:24 +0000 (14:33 +0000)]
Updated Slovak translation

9 years agoexamples: use G_DECLARE_FINAL_TYPE in applications
Mohammed Sadiq [Tue, 6 Sep 2016 17:56:22 +0000 (23:26 +0530)]
examples: use G_DECLARE_FINAL_TYPE in applications

G_DECLARE_FINAL_TYPE was introduced in glib 2.44.
We shall use that now so that lots of boilerplate code can
be reduced.

https://bugzilla.gnome.org/show_bug.cgi?id=770278

9 years agoUpdated Galician translations
Fran Dieguez [Fri, 9 Sep 2016 23:10:49 +0000 (01:10 +0200)]
Updated Galician translations

9 years agoplacesview: keep reference during network fetching
Christian Kellner [Mon, 5 Sep 2016 11:41:16 +0000 (13:41 +0200)]
placesview: keep reference during network fetching

Analogous to (un)mount operation, we now keep a reference around
during the ongoing operation and make use of the destroyed flag
to check if we are still alive or if we have been cancelled as
a result of the widget being destroyed.

https://bugzilla.gnome.org/show_bug.cgi?id=764979

9 years agoplacesview: override destory to cancel ongoing ops
Christian Kellner [Mon, 5 Sep 2016 11:19:29 +0000 (13:19 +0200)]
placesview: override destory to cancel ongoing ops

Since we hold on to a reference during (un)mount operations, we
don't trigger the cancellation of operations in finalize anymore.
Instead we now override the GtkWidget's destroy() and cancel any
ongoing operations there.

https://bugzilla.gnome.org/show_bug.cgi?id=764979

9 years agoplacesview: properly recover from cancellation
Christian Kellner [Sun, 4 Sep 2016 13:28:11 +0000 (15:28 +0200)]
placesview: properly recover from cancellation

The current code wrongly assumes that cancellation can only happen
as a result widget finalization, and consequentially does not
properly recover from it. Therefore if the operation is cancelled
as a result of user interaction, the entry is will stay disabled
and the spinner will keep spinning. This is fixed by removal of
the early bail out in case of cancellation.

https://bugzilla.gnome.org/show_bug.cgi?id=764979

9 years agoplacesview: hold onto ref during (un)mount ops
Christian Kellner [Sun, 4 Sep 2016 09:47:55 +0000 (11:47 +0200)]
placesview: hold onto ref during (un)mount ops

During mount and unmount opertions we keep a reference to the
GtkPlacesView around, so we have a valid view for the callback
code, even in the case that othe external references have been
dropped (i.e. the containing window gets destroyed).

https://bugzilla.gnome.org/show_bug.cgi?id=764979

9 years agogdkscreen-x11: Enable RandR on VirtualBox
Takao Fujiwara [Fri, 9 Sep 2016 16:54:13 +0000 (01:54 +0900)]
gdkscreen-x11: Enable RandR on VirtualBox

RandR 1.5 is enabled on VirtualBox guest of Fedora 25 but
XRROutputInfo->name is "default". If init_randr15() does not
return TRUE, the monitor size sets 0 because gdk_screen_get_width()
returns 0.

This problem causes GtkStatusIcon not to show the activate menu.

https://bugzilla.gnome.org/show_bug.cgi?id=771033

9 years agoUpdated Portuguese translation
Tiago Santos [Fri, 9 Sep 2016 16:59:19 +0000 (16:59 +0000)]
Updated Portuguese translation

9 years ago[l10n] update Persian translations
Arash Mousavi [Fri, 9 Sep 2016 13:56:04 +0000 (18:26 +0430)]
[l10n] update Persian translations

9 years agoUpdate Galician translations
Fran Dieguez [Thu, 8 Sep 2016 21:03:03 +0000 (23:03 +0200)]
Update Galician translations

Signed-off-by: Fran Dieguez <fran.dieguez@mabishu.com>