gtk4.git
11 years agoevent-controller: Chain up on dispose
Carlos Garnacho [Wed, 14 Jan 2015 15:03:10 +0000 (16:03 +0100)]
event-controller: Chain up on dispose

11 years agoUpdated Russian translation
Stas Solovey [Wed, 14 Jan 2015 12:05:57 +0000 (12:05 +0000)]
Updated Russian translation

11 years agogtkplacessidebar: insert bookmark by dnd in correct position
Nelson Benítez León [Sat, 10 Jan 2015 20:11:14 +0000 (01:11 +0500)]
gtkplacessidebar: insert bookmark by dnd in correct position

Add needed COLUMN INDEX to "new bookmark" row based on previous
bookmark index plus one.

Fixes bug 742727

11 years agoFix accessible_class->widget_set initialization
Maks Naumov [Sun, 11 Jan 2015 21:34:38 +0000 (13:34 -0800)]
Fix accessible_class->widget_set initialization

Signed-off-by: Maks Naumov <maksqwe1@ukr.net>
https://bugzilla.gnome.org/show_bug.cgi?id=742778

11 years agoUpdated Galician translations
Fran Dieguez [Mon, 12 Jan 2015 21:53:18 +0000 (22:53 +0100)]
Updated Galician translations

11 years agoUpdated Spanish translation
Daniel Mustieles [Mon, 12 Jan 2015 11:20:26 +0000 (12:20 +0100)]
Updated Spanish translation

11 years agoProperly check result of g_utf8_get_char_validated()
Maks Naumov [Sun, 11 Jan 2015 19:27:37 +0000 (11:27 -0800)]
Properly check result of g_utf8_get_char_validated()

g_utf8_get_char_validated() may return -1 or -2
return type is gunichar(guint32)

Therefore such checks like 'gunichar < 0' or 'gunichar > 0'
are always 'false' or 'true'(except when gunichar == 0).

Signed-off-by: Maks Naumov <maksqwe1@ukr.net>
https://bugzilla.gnome.org/show_bug.cgi?id=742774

11 years agoFix check for 'y' padding in gtk_cell_renderer_set_padding()
Maks Naumov [Sun, 11 Jan 2015 15:01:40 +0000 (07:01 -0800)]
Fix check for 'y' padding in gtk_cell_renderer_set_padding()

Signed-off-by: Maks Naumov <maksqwe1@ukr.net>
https://bugzilla.gnome.org/show_bug.cgi?id=742764

11 years agoRemove an unused define
Matthias Clasen [Sun, 11 Jan 2015 20:14:55 +0000 (15:14 -0500)]
Remove an unused define

11 years agoProperly ignore BadWindow and BadDrawable in get_child_info_handler()
Maks Naumov [Sun, 11 Jan 2015 18:34:56 +0000 (10:34 -0800)]
Properly ignore BadWindow and BadDrawable in get_child_info_handler()

Signed-off-by: Maks Naumov <maksqwe1@ukr.net>
https://bugzilla.gnome.org/show_bug.cgi?id=742771

11 years agoForgotten file
Matthias Clasen [Sun, 11 Jan 2015 02:16:21 +0000 (21:16 -0500)]
Forgotten file

11 years agoHighContrast: Don't set nonexisting style properties
Matthias Clasen [Sun, 11 Jan 2015 02:14:56 +0000 (21:14 -0500)]
HighContrast: Don't set nonexisting style properties

These don't exist, so setting them can't do any good.

11 years agoAdwaita: Don't set nonexisting style properties
Matthias Clasen [Sun, 11 Jan 2015 02:14:02 +0000 (21:14 -0500)]
Adwaita: Don't set nonexisting style properties

These don't exist, so setting them can't do any good.

11 years agoHighContrast: Also fix separators in file chooser buttons
Matthias Clasen [Sun, 11 Jan 2015 02:12:08 +0000 (21:12 -0500)]
HighContrast: Also fix separators in file chooser buttons

Same fix as for combo box separators.

11 years agoAdwaita: Also fix separators in file chooser buttons
Matthias Clasen [Sun, 11 Jan 2015 02:09:30 +0000 (21:09 -0500)]
Adwaita: Also fix separators in file chooser buttons

Same fix as for combo box separators.

11 years agogtkfilechooserwidget: When using an extra widget, show or hide the filter combo as...
Iain Lane [Fri, 9 Jan 2015 15:31:08 +0000 (15:31 +0000)]
gtkfilechooserwidget: When using an extra widget, show or hide the filter combo as appropriate

https://bugzilla.gnome.org/show_bug.cgi?id=742653

11 years agoHighContrast: draw combo box separators
Matthias Clasen [Sun, 11 Jan 2015 01:28:45 +0000 (20:28 -0500)]
HighContrast: draw combo box separators

Same fix as the previous commit.

11 years agoAdwaita: draw combo box separators
Matthias Clasen [Sun, 11 Jan 2015 01:25:56 +0000 (20:25 -0500)]
Adwaita: draw combo box separators

Killing the separator in the button had the unintended side effect
of also killing separators in the popup.

https://bugzilla.gnome.org/show_bug.cgi?id=742704

11 years agowidget-factory: Add a combo box separator example
Matthias Clasen [Sun, 11 Jan 2015 01:14:24 +0000 (20:14 -0500)]
widget-factory: Add a combo box separator example

This shows that combo box separators are currently broken
in Adwaita.

11 years agoGtkTextView: allow to extend text selection with mouse scrollwheel
Nelson Benitez Leon [Tue, 22 Oct 2013 10:24:55 +0000 (12:24 +0200)]
GtkTextView: allow to extend text selection with mouse scrollwheel

Fixes bug 710612

11 years agofile chooser: Add some padidng
Matthias Clasen [Sat, 10 Jan 2015 19:27:52 +0000 (14:27 -0500)]
file chooser: Add some padidng

https://bugzilla.gnome.org/show_bug.cgi?id=742631

11 years agoForgotten files
Matthias Clasen [Sat, 10 Jan 2015 19:35:45 +0000 (14:35 -0500)]
Forgotten files

11 years agoAdwaita: Fix treeview separators in backdrop
Matthias Clasen [Sat, 10 Jan 2015 18:04:10 +0000 (13:04 -0500)]
Adwaita: Fix treeview separators in backdrop

https://bugzilla.gnome.org/show_bug.cgi?id=742702

11 years agostylecontext: Stop animations on manual invalidation
Benjamin Otte [Sat, 10 Jan 2015 12:58:39 +0000 (13:58 +0100)]
stylecontext: Stop animations on manual invalidation

If a gtk_style_context_invalidate() is called on a widget's style
context (which nobody should ever do, sheesh!) and we're animating, stop
the animations.

Fixes crashers in Nautilus.

11 years agoGtkScrolledWindow: Don't start fading right away
Matthias Clasen [Sat, 10 Jan 2015 02:13:10 +0000 (21:13 -0500)]
GtkScrolledWindow: Don't start fading right away

When we get a leave event, we need to remove the over bit,
since we may not get motion events that tell us when we lose
proximity, but we don't have to fade out right away - that
will happen on its on due to inactivity a bit later. Having
the indicator stay visible for a bit feels more relaxed.

11 years agoGtkScrolledWindow: Factor out a constant
Matthias Clasen [Sat, 10 Jan 2015 02:12:00 +0000 (21:12 -0500)]
GtkScrolledWindow: Factor out a constant

Define INDICATOR_CLOSE_DISTANCE instead of hardcoding the same
number of pixels in 4 different places.

11 years agoGtkScrolledWindow: Don't require proximity for track points
Matthias Clasen [Sat, 10 Jan 2015 02:09:22 +0000 (21:09 -0500)]
GtkScrolledWindow: Don't require proximity for track points

These are hard to navigate precisely, so show wide scrollbars
right away.

11 years agoGtkScrolledWindow: Drop device manager complications
Matthias Clasen [Sat, 10 Jan 2015 01:52:03 +0000 (20:52 -0500)]
GtkScrolledWindow: Drop device manager complications

We no longer use the presence or absence of particular device
types to determine which scrolling to use.

11 years agoGtkScrolledWindow: Drop a redundant field
Matthias Clasen [Sat, 10 Jan 2015 01:33:39 +0000 (20:33 -0500)]
GtkScrolledWindow: Drop a redundant field

The Indicator.enabled field was not carrying any extra information,
so just use priv->use_indicators instead.

11 years agoGtkScrolledWindow: Rename a private field
Matthias Clasen [Sat, 10 Jan 2015 01:26:07 +0000 (20:26 -0500)]
GtkScrolledWindow: Rename a private field

touch_mode -> use_indicators, more to the point.

11 years agox11/gl: Do not specify the GLX context version
Emmanuele Bassi [Sat, 10 Jan 2015 00:07:51 +0000 (00:07 +0000)]
x11/gl: Do not specify the GLX context version

If we use GDK_GL_PROFILE_3_2_CORE we are asking for a core profile
according to the GLX_ARB_create_context_profile extension. For that,
we pass the GLX_CONTEXT_CORE_PROFILE_BIT_ARB value for the
GLX_CONTEXT_PROFILE_MASK_ARB attribute.

The specification for the extension says that:

  If the requested OpenGL version is less than 3.2,
  GLX_CONTEXT_PROFILE_MASK_ARB is ignored and the functionality
  of the context is determined solely by the requested version.

Since we're asking for a core profile, we assume a GL version greater
than or equal to 3.2; thus, we don't need to specify the
GLX_CONTEXT_MAJOR_VERSION_ARB or the GLX_CONTEXT_MINOR_VERSION_ARB
attributes, and instead just rely on whatever version GLX gives us.

This seems to work around a strange issue in Mesa; if we ask for a core
profile and any version > 3.0, we get broken rendering on any shared
context we create.

11 years agoinspector: Don't use GSettings directly
Matthias Clasen [Fri, 9 Jan 2015 20:28:53 +0000 (15:28 -0500)]
inspector: Don't use GSettings directly

It is not necessary here, and using GtkSettings gives us
a greater chance to not fail e.g. on Windows.

https://bugzilla.gnome.org/show_bug.cgi?id=742664

11 years agoinspector: be careful about gsize vs guint64
Matthias Clasen [Fri, 9 Jan 2015 19:44:26 +0000 (14:44 -0500)]
inspector: be careful about gsize vs guint64

Since gsize doesn't work as a typename in a ui file,
we have to be careful not to pass pointers to wrongly
sized variables when getting the guint64 values out
of the model.

https://bugzilla.gnome.org/show_bug.cgi?id=742664

11 years agodocs: Add missing quote
Emmanuele Bassi [Fri, 9 Jan 2015 16:42:55 +0000 (16:42 +0000)]
docs: Add missing quote

The missing " breaks the syntax highlighting of the GtkImageMenuItem
example.

11 years agodocs: Add Since for GtkCellRendererPixbuf:surface
David King [Fri, 9 Jan 2015 14:37:17 +0000 (14:37 +0000)]
docs: Add Since for GtkCellRendererPixbuf:surface

11 years agodocs: Fix some old live.gnome.org URLs
David King [Tue, 8 Apr 2014 15:42:36 +0000 (16:42 +0100)]
docs: Fix some old live.gnome.org URLs

11 years agoUpdated POTFILES.skip
Piotr Drąg [Fri, 9 Jan 2015 13:16:03 +0000 (14:16 +0100)]
Updated POTFILES.skip

11 years agoFix GTK_OVERLAY_SCROLLING=0
Matthias Clasen [Fri, 9 Jan 2015 11:45:51 +0000 (06:45 -0500)]
Fix GTK_OVERLAY_SCROLLING=0

This was broken by the recent more dynamic approach.

11 years agoFix comment typo
Matthias Clasen [Fri, 9 Jan 2015 11:21:37 +0000 (06:21 -0500)]
Fix comment typo

11 years agoinspector: Show object address
Matthias Clasen [Fri, 9 Jan 2015 05:07:01 +0000 (00:07 -0500)]
inspector: Show object address

This is useful information when running the application
in a debugger.
https://bugzilla.gnome.org/show_bug.cgi?id=742518

11 years agoPoint to copy/cut/delete named icons in the stock item deprecations
Emmanuele Bassi [Fri, 26 Dec 2014 16:04:26 +0000 (16:04 +0000)]
Point to copy/cut/delete named icons in the stock item deprecations

We still use symbolic icons for Copy, Cut, and Delete operations inside
toolbars and some menus, so it's worth pointing to the symbolic icon
names in the deprecation notices.

https://bugzilla.gnome.org/show_bug.cgi?id=741969

11 years agofile chooser: Don't force a scrollbar
Matthias Clasen [Fri, 9 Jan 2015 03:27:47 +0000 (22:27 -0500)]
file chooser: Don't force a scrollbar

No point in showing the scrollbar when it is not needed.
https://bugzilla.gnome.org/show_bug.cgi?id=742343

11 years agogtklabel: Don't accept unhandled mouse buttons
Timm Bäder [Fri, 26 Dec 2014 22:04:37 +0000 (23:04 +0100)]
gtklabel: Don't accept unhandled mouse buttons

If the pressed mouse button neither triggers a context menu nor
activates a link, set the sequence's state to DENIED.

https://bugzilla.gnome.org/show_bug.cgi?id=742010

11 years agocssimage: Fix angle computation for gradients to corners
Benjamin Otte [Thu, 8 Jan 2015 18:30:19 +0000 (19:30 +0100)]
cssimage: Fix angle computation for gradients to corners

11 years agoUpdated Basque language
Inaki Larranaga Murgoitio [Thu, 8 Jan 2015 17:32:00 +0000 (18:32 +0100)]
Updated Basque language

11 years agoUpdated Basque language
Inaki Larranaga Murgoitio [Thu, 8 Jan 2015 14:41:15 +0000 (15:41 +0100)]
Updated Basque language

11 years agostylecontext: Handle theme changes for cached styles
Benjamin Otte [Wed, 7 Jan 2015 17:06:59 +0000 (18:06 +0100)]
stylecontext: Handle theme changes for cached styles

We need to clear the cache manually on full revalidates because
_gtk_css_change_for_child() will clear the full revalidation flags.
And then gtk_style_context_update_cache() will not do the right thing
(which is to clear itself).

11 years agostylecontext: Rename GtkStyleInfo to GtkCssNode
Benjamin Otte [Wed, 7 Jan 2015 14:44:53 +0000 (15:44 +0100)]
stylecontext: Rename GtkStyleInfo to GtkCssNode

11 years agostylecontext: update_animating() properly
Benjamin Otte [Tue, 6 Jan 2015 11:08:49 +0000 (12:08 +0100)]
stylecontext: update_animating() properly

Because we can switch from animating to non-animating pretty much
anywhere, do the check for animations unconditionally instead of trying
to cram it into the correct if path (and failing).

11 years agostack: Remove commented out line
Benjamin Otte [Tue, 6 Jan 2015 11:08:17 +0000 (12:08 +0100)]
stack: Remove commented out line

It always shows up in my greps and confuses me.

11 years agostylecontext: Add a parent style cache
Benjamin Otte [Mon, 5 Jan 2015 23:38:54 +0000 (00:38 +0100)]
stylecontext: Add a parent style cache

We now cache the results of lookups on the parent GtkCssStyle. This
allows sharing styles between widgets (recursively). However, this
only works if the styles can't potentially depend on siblings -
neither directly via sibling selectors or via position pseudo-classes
like :first-child.
Unfortunately, Adwaita currently uses first-child a lot, and in
particular for labels, which are the most common widgets.

The big benefits of this change are both less CPU - due to not needing
to compute styles again - and less memory usage - due to sharing of
the styles between widgets. Here's some nonscientific numbers I
collected while pondering the usefulness of this patch:

                    glade   glade   widget
                    demo    demo    factory
                    runtime styles  styles
Adwaita before      19.1s   5800    1150
Adwaita now         18.9s   3800     970
Adwaita hacked now  14.5s   3100     910
simple before       11.3s   5800    1150
simple now          10.8s   1300     590

Adwaita: Adwaita as provided by GTK
Adwaita hacked: Adwaita with the first/last-child for GtkLabel removed
simple: A 250 lines simple GTK theme I use for testing
before: This patch not applied
now: this patch applied

glade demo runtime: Starting glade opening a large file and closing it
glade demo styles: GtkCssStaticStyle objects after opening glade with
                   the large file as per inspector
widget factory styles: GtkCssStaticStyle objects after startup as per
                       inspector

11 years agocssstyle: Have a default style
Benjamin Otte [Mon, 5 Jan 2015 15:13:18 +0000 (16:13 +0100)]
cssstyle: Have a default style

This way we can initialize the default style info with a style and
ddon't need to special case NULL everywhere.

11 years agostylecontext: Use correct parent on updates
Benjamin Otte [Mon, 5 Jan 2015 04:57:45 +0000 (05:57 +0100)]
stylecontext: Use correct parent on updates

Previously we looked at the save/restore state when determining the
parent. This is wrong in the case where we're updating the cache though.

So we now save the parent in the style info.

11 years agostylecontext: Fix cache updating
Benjamin Otte [Tue, 23 Dec 2014 07:19:39 +0000 (08:19 +0100)]
stylecontext: Fix cache updating

After the parent changes in commit
3a337156d11a86c7a88f1f30a09276fdf6f63008 we need to refresh the cached
styles after the current style. After all, they now depend on the base
style.

11 years agocssstaticstyle: Handle relevant change
Benjamin Otte [Tue, 23 Dec 2014 07:10:30 +0000 (08:10 +0100)]
cssstaticstyle: Handle relevant change

Relevant change is returned from style lookups, so keep is with the
object that holds those values.

11 years agostylecontext: Use correct parent style for lookups
Benjamin Otte [Sun, 21 Dec 2014 20:30:51 +0000 (21:30 +0100)]
stylecontext: Use correct parent style for lookups

After 3a337156d11a86c7a88f1f30a09276fdf6f63008 style lookups still used
the parent context's style as the parent style, even though after a
gtk_style_context_save() the root style of the style context is the
proper parent.

Testcase attached.

11 years agocssnodedeclaration: Store the GType of the node
Benjamin Otte [Sun, 21 Dec 2014 19:26:26 +0000 (20:26 +0100)]
cssnodedeclaration: Store the GType of the node

11 years agocssstyle: Move function from vfunc to static func
Benjamin Otte [Wed, 17 Dec 2014 12:30:51 +0000 (13:30 +0100)]
cssstyle: Move function from vfunc to static func

compute_dependencies() is only used internally by GtkCssStaticStyle,
so there's no need to have it elsewhere.

11 years agocssstaticstyle: Move function
Benjamin Otte [Wed, 17 Dec 2014 12:22:03 +0000 (13:22 +0100)]
cssstaticstyle: Move function

And with that move, GtkCssStaticStyle is immutable.

11 years agocssstaticstyle: Move function
Benjamin Otte [Wed, 17 Dec 2014 04:52:13 +0000 (05:52 +0100)]
cssstaticstyle: Move function

We want the new() return values to be immutable, so we have to move the
code that modifes them.

11 years agostylecontext: Refactor update_properties()
Benjamin Otte [Wed, 17 Dec 2014 03:18:44 +0000 (04:18 +0100)]
stylecontext: Refactor update_properties()

It now always returns a new instance.

11 years agostylecontext: Simplify function
Benjamin Otte [Tue, 16 Dec 2014 05:47:08 +0000 (06:47 +0100)]
stylecontext: Simplify function

Instead of computing the changes manually everywhere from the parent
changes, just pass in the parent changes.

11 years agostylecontext: Simplify code
Benjamin Otte [Tue, 16 Dec 2014 05:40:10 +0000 (06:40 +0100)]
stylecontext: Simplify code

Compute the differences only once and not in both if branches.

11 years agocssanimatedstyle: Apply animation values after creation
Benjamin Otte [Tue, 16 Dec 2014 04:57:36 +0000 (05:57 +0100)]
cssanimatedstyle: Apply animation values after creation

This makes GtkCssAnimatedStyle immutable.

11 years agocssanimatedstyle: Make advancing the style return a new style
Benjamin Otte [Tue, 16 Dec 2014 03:15:25 +0000 (04:15 +0100)]
cssanimatedstyle: Make advancing the style return a new style

This is another step towards making CssStyles immutable.

11 years agostylecontext: Fold a function into only caller
Benjamin Otte [Mon, 15 Dec 2014 13:22:00 +0000 (14:22 +0100)]
stylecontext: Fold a function into only caller

11 years agocssanimatedstyle: Avoid creating animation
Benjamin Otte [Mon, 15 Dec 2014 03:17:42 +0000 (04:17 +0100)]
cssanimatedstyle: Avoid creating animation

... when no animation exists.

This makes the function name kinda wrong, but I'm not sure what else to
call it.

11 years agocssanimatedstyle: Remove unused function
Benjamin Otte [Mon, 15 Dec 2014 01:54:35 +0000 (02:54 +0100)]
cssanimatedstyle: Remove unused function

11 years agocssanimatedstyle: Merge two functions
Benjamin Otte [Mon, 15 Dec 2014 00:58:11 +0000 (01:58 +0100)]
cssanimatedstyle: Merge two functions

11 years agostylecontext: Make build_properties create the style
Benjamin Otte [Sun, 14 Dec 2014 22:20:19 +0000 (23:20 +0100)]
stylecontext: Make build_properties create the style

11 years agostylecontext: Make static styles static
Benjamin Otte [Sun, 14 Dec 2014 16:06:27 +0000 (17:06 +0100)]
stylecontext: Make static styles static

Instead of keeping an animated style everywhere, we replace it with the
static style when nothing gets animated.

Apart from making the code cleaner, this gets rid of a bunch of animated
style values that do nothing but wrap a static style.

11 years agostylecontext: Split function
Benjamin Otte [Sun, 14 Dec 2014 14:46:18 +0000 (15:46 +0100)]
stylecontext: Split function

11 years agostylecontext: Only look up stuff in the cache that can be in the cache
Benjamin Otte [Sat, 13 Dec 2014 01:07:34 +0000 (02:07 +0100)]
stylecontext: Only look up stuff in the cache that can be in the cache

Only saved styles are stored in the cache so we should look only those
up there.

11 years agostylecontext: don't unset the current style info when clearing cache
Benjamin Otte [Mon, 27 Oct 2014 03:22:49 +0000 (04:22 +0100)]
stylecontext: don't unset the current style info when clearing cache

Instead, do it manually.

11 years agostylecontext: Always pass the static style
Benjamin Otte [Sun, 26 Oct 2014 23:25:18 +0000 (00:25 +0100)]
stylecontext: Always pass the static style

... to build_properties.
There is always one available, so we can insist on that one.

This simplifies a bunch of code.

11 years agostylecontext: Use a static style when saved
Benjamin Otte [Sun, 26 Oct 2014 18:38:38 +0000 (19:38 +0100)]
stylecontext: Use a static style when saved

The only style that is animated is the style of the unsaved primary
node. So there's no need to create animated style objects for the other
ones.

There is a bunch of ugliness in the code currently. Further commits are
expected to fix them.

11 years agocssstyle: Add GtkCssStaticStyle
Benjamin Otte [Sun, 26 Oct 2014 04:07:06 +0000 (05:07 +0100)]
cssstyle: Add GtkCssStaticStyle

For now, this is only an implementation detail of the animated style.

The idea is to use GtkCssStaticStyle as the result of CSS queries and
then put a GtkCssAnimatedStyle on top that manages the animations. The
neat thing about this is that you can cache the static values.

11 years agocssstyle: Split into GtkCssStyle and GtkCssAnimatedStyle
Benjamin Otte [Thu, 23 Oct 2014 02:10:03 +0000 (04:10 +0200)]
cssstyle: Split into GtkCssStyle and GtkCssAnimatedStyle

GtkCssStyle is the base class to be used for all types of styles that do
exist.

GtkCssAnimatedStyle is the only implementation so far, that is exactly a
copy/paste of the old GtkCssStyle code.

11 years agocssstyle: Rename GtkCssComputedValues => GtkCssStyle
Benjamin Otte [Wed, 22 Oct 2014 21:41:50 +0000 (23:41 +0200)]
cssstyle: Rename GtkCssComputedValues => GtkCssStyle

This is literally just renaming of the object (and the associated source
files). No other changes are in there.

11 years agocssstylefuncs: Rename functions
Benjamin Otte [Wed, 22 Oct 2014 20:36:38 +0000 (22:36 +0200)]
cssstylefuncs: Rename functions

This is in preaparation for the big GtkCssComputedValues => GtkCssStyle
rename

11 years agoUpdated German translation
Christian Kirbach [Tue, 6 Jan 2015 22:42:35 +0000 (22:42 +0000)]
Updated German translation

11 years agoUpdated British English translation
Bruce Cowan [Mon, 5 Jan 2015 20:40:52 +0000 (20:40 +0000)]
Updated British English translation

11 years agocssselector: Reintroduce GOT_MATCH
Benjamin Otte [Mon, 5 Jan 2015 20:15:19 +0000 (21:15 +0100)]
cssselector: Reintroduce GOT_MATCH

I removed it in 14f5ce710856708dc80fcc98cb5eb2ba257f34ad because I
thought it was unnecessary, but it wasn't. When we build a tree like
this:

.matches ─┬─ .doesntmatch
          └─ .alsodoesntmatch

We would get the changes for the .matches part returned. This is however
only right if that node of the tree contains results. If results only
exist with the child nodes (all of which don't match), then this part
should not match either.

11 years agocssselector: The if was the wrong way around
Benjamin Otte [Mon, 5 Jan 2015 20:00:36 +0000 (21:00 +0100)]
cssselector: The if was the wrong way around

We want to keep matching as long as selectors are simple. Otherwise we'd
never be matching, because the first selector is always simple. Oops.

11 years agocellrenderertext: notify inverted
Sebastien Lafargue [Sun, 4 Jan 2015 11:40:27 +0000 (12:40 +0100)]
cellrenderertext: notify inverted

The notify for "size" and "size-points" properties
are inverted.

https://bugzilla.gnome.org/show_bug.cgi?id=742311

11 years agocalendar: Handle invalid dates
Emmanuele Bassi [Fri, 2 Jan 2015 23:05:29 +0000 (23:05 +0000)]
calendar: Handle invalid dates

GtkCalendar can have an invalid date — mostly at initialization. This
means that GDateTime construction may fail. We need to handle that case
gracefully, like the old code did.

This fixes the `notify` test suite, which started failing with:

/Notification/GtkCalendar:
GLib-CRITICAL **: g_date_time_get_day_of_week: assertion 'datetime != NULL' failed

inside the Continuous builder.

11 years agotextview: fix duplicated binding entries. use kp variants instead.
Christian Hergert [Thu, 1 Jan 2015 07:27:55 +0000 (23:27 -0800)]
textview: fix duplicated binding entries. use kp variants instead.

This appears to have accidentally used the normal keypresses twice instead
of the KP variants. This resulted in shift+insert pasting twice.

https://bugzilla.gnome.org/show_bug.cgi?id=742170

11 years agoUpdated Basque language
Inaki Larranaga Murgoitio [Wed, 31 Dec 2014 16:46:21 +0000 (17:46 +0100)]
Updated Basque language

11 years agoUpdated Hungarian translation
Balázs Úr [Wed, 31 Dec 2014 13:29:14 +0000 (13:29 +0000)]
Updated Hungarian translation

11 years agoUpdated Hungarian translation
Gábor Kelemen [Wed, 31 Dec 2014 13:02:56 +0000 (13:02 +0000)]
Updated Hungarian translation

11 years agoUpdated Hungarian translation
Balázs Úr [Wed, 31 Dec 2014 12:03:47 +0000 (12:03 +0000)]
Updated Hungarian translation

11 years agoUpdated Vietnamese translation
Trần Ngọc Quân [Wed, 31 Dec 2014 08:50:54 +0000 (15:50 +0700)]
Updated Vietnamese translation

Signed-off-by: Trần Ngọc Quân <vnwildman@gmail.com>
11 years agoFix typos in documentation
Ignacio Casal Quinteiro [Tue, 30 Dec 2014 12:46:11 +0000 (13:46 +0100)]
Fix typos in documentation

11 years agogtkstylecontext: Don't try to emit a signal when finalizing
Jasper St. Pierre [Mon, 29 Dec 2014 02:11:49 +0000 (18:11 -0800)]
gtkstylecontext: Don't try to emit a signal when finalizing

11 years agogdkeventsource-wayland: Fix style
Jasper St. Pierre [Mon, 29 Dec 2014 01:54:09 +0000 (17:54 -0800)]
gdkeventsource-wayland: Fix style

11 years agoinspector: Update translations .ui.h file
Jasper St. Pierre [Mon, 29 Dec 2014 02:13:24 +0000 (18:13 -0800)]
inspector: Update translations .ui.h file

11 years agoInspector: Show some X display characteristics
Matthias Clasen [Sun, 28 Dec 2014 16:36:10 +0000 (11:36 -0500)]
Inspector: Show some X display characteristics

This helps diagnose e.g. csd problems.

11 years agoRestore a missing file
Matthias Clasen [Sun, 28 Dec 2014 16:06:24 +0000 (11:06 -0500)]
Restore a missing file

No wonder nobody has tried my awesome magnifier yet...

11 years agomir: Add missing reference to gdkmir-private.h
Rico Tzschichholz [Sun, 28 Dec 2014 12:34:16 +0000 (13:34 +0100)]
mir: Add missing reference to gdkmir-private.h

11 years agoUpdated vi translation for gtk+properties
Trần Ngọc Quân [Sun, 28 Dec 2014 02:20:14 +0000 (09:20 +0700)]
Updated vi translation for gtk+properties

Signed-off-by: Trần Ngọc Quân <vnwildman@gmail.com>