]> Pileus Git - ~andy/gtk/log
~andy/gtk
12 years agogdk: Make gen-keyname-table.pl work again
Bastien Nocera [Tue, 25 Oct 2011 13:34:27 +0000 (14:34 +0100)]
gdk: Make gen-keyname-table.pl work again

keyname-table.h was modified by hand, and gen-keyname-table.pl
couldn't be used to generate that file again.

https://bugzilla.gnome.org/show_bug.cgi?id=662628

12 years agoUpdate grid tests to match
Matthias Clasen [Wed, 26 Oct 2011 03:20:37 +0000 (23:20 -0400)]
Update grid tests to match

The previous fix changed some test results here.

12 years agoFix gtk_grid_attach_next_to
Matthias Clasen [Wed, 26 Oct 2011 03:18:21 +0000 (23:18 -0400)]
Fix gtk_grid_attach_next_to

The function did not behave as intended when @sibling is NULL.
Also add an example to the docs to clarify the intended behaviour.

12 years agogtkrc.key.mac: add Command-cursor text navigation
Michael Natterer [Tue, 25 Oct 2011 20:19:11 +0000 (22:19 +0200)]
gtkrc.key.mac: add Command-cursor text navigation

and some emacs-ish Control bindings that work in native widgets.
Patch from Michael Hutchinson.
(cherry picked from commit ccf12f7b406ecbd8f0c26b0e6dc86d4593144dab)

12 years agogtkrc.key.mac: remove half the file, it was an accidential double paste
Michael Natterer [Tue, 25 Oct 2011 15:01:25 +0000 (17:01 +0200)]
gtkrc.key.mac: remove half the file, it was an accidential double paste
(cherry picked from commit 79a92f99a9dbdc7b1a651b0e8665807bd89c6632)

12 years agoPost-release version bump
Matthias Clasen [Tue, 25 Oct 2011 14:22:15 +0000 (10:22 -0400)]
Post-release version bump

Halfway to 666 !

12 years agoBump version to 3.3.2
Matthias Clasen [Tue, 25 Oct 2011 13:15:50 +0000 (09:15 -0400)]
Bump version to 3.3.2

12 years agoUpdates
Matthias Clasen [Tue, 25 Oct 2011 11:37:32 +0000 (07:37 -0400)]
Updates

12 years agoFix more test breakage
Matthias Clasen [Fri, 14 Oct 2011 19:52:24 +0000 (15:52 -0400)]
Fix more test breakage

12 years agoa11y: Hack around infinite loops in parent setting
Benjamin Otte [Mon, 3 Oct 2011 15:05:40 +0000 (17:05 +0200)]
a11y: Hack around infinite loops in parent setting

This is kind of a hack to get rid of infinite loops that occur when
child accessibles try to set their parent upon creation but the parent
accessible creates its children in the initialize vfunc. Because in that
case, the parent will not have an accessible set when the child tries to
access it, because it is still initializing itself. Which will cause a
new accessible to be created.

https://bugzilla.gnome.org/show_bug.cgi?id=660687

12 years agoFix make check
Matthias Clasen [Fri, 14 Oct 2011 15:43:40 +0000 (11:43 -0400)]
Fix make check

Somebody committed a behaviour change that broke his own
tests...tsk, tsk.

12 years agoAdd a css parser test for transparent
Matthias Clasen [Fri, 14 Oct 2011 15:23:43 +0000 (11:23 -0400)]
Add a css parser test for transparent

12 years agoprintunixdialog: fix a GtkTable->GtkGrid typo
Cosimo Cecchi [Mon, 24 Oct 2011 20:31:50 +0000 (16:31 -0400)]
printunixdialog: fix a GtkTable->GtkGrid typo

Code was treating the GtkScrolledWindow as a grid, not the grid itself.

12 years agoFix handling of GtkCellRendererProgress::inverted
Matthias Clasen [Mon, 24 Oct 2011 17:50:00 +0000 (13:50 -0400)]
Fix handling of GtkCellRendererProgress::inverted

We were setting the wrong field.
https://bugzilla.redhat.com/show_bug.cgi?id=747761

12 years agonotebook: update tab area allocation when switching page
Cosimo Cecchi [Mon, 24 Oct 2011 17:12:21 +0000 (13:12 -0400)]
notebook: update tab area allocation when switching page

When the active page is switched, it's important
gtk_notebook_pages_allocate() is called after the bulk of
gtk_notebook_real_switch_page() is run, as the former allocates a
different space according to the currently active tab, which is set by
the latter.
gtk_notebook_pages_allocate() already calls gtk_notebook_redraw_tabs()
when the allocation changes, so just move its call down to
gtk_notebook_real_switch_page() to fix the bug.

https://bugzilla.gnome.org/show_bug.cgi?id=662629

12 years agowin32: fix unterminated #ifdef error
Dieter Verfaillie [Sun, 23 Oct 2011 19:43:16 +0000 (21:43 +0200)]
win32: fix unterminated #ifdef error

By removing all ifdefs as gtk/gtkwin32.c is only built
on win32 anyway.

https://bugzilla.gnome.org/show_bug.cgi?id=662481

12 years agotooltip: only update shape if the window is realized
Rui Matos [Sat, 22 Oct 2011 16:38:29 +0000 (17:38 +0100)]
tooltip: only update shape if the window is realized

If we get a composited-changed signal before the tooltip is ever realized we
have nothing to do.

This actually prevents a segfault in maybe_update_shape() because calling
gtk_widget_get_window() on a non-realized widget returns NULL.

https://bugzilla.gnome.org/show_bug.cgi?id=662467

12 years agogtk: move _gtk_modules_has_mixed_deps() to gtkmodlesprivate.h
Michael Natterer [Sun, 23 Oct 2011 11:57:07 +0000 (13:57 +0200)]
gtk: move _gtk_modules_has_mixed_deps() to gtkmodlesprivate.h

and remove gtkmainprivate.h completely.

12 years agogtk: clean up the private horror
Michael Natterer [Sat, 22 Oct 2011 06:48:13 +0000 (08:48 +0200)]
gtk: clean up the private horror

- add gtkmodulesprivate.h and move stuff there from gtkprivate.h
- add gtkprivate.c and move stuff there from gtkmain.c
- add gtkwin32.c and move stuff there from gtkmain.c
- don't redefine GTK_DATADIR and friends in gtkprivate.h
- have _gtk_get_datadir() and friends on all platforms
- remove the horrid hacks where gtkprivate.h can't be included,
  or must be included later due to redefinition of the compile-time
  directories

12 years agogtkquartz: don't free the string returned by get_bundle_path()
Michael Natterer [Sat, 22 Oct 2011 21:08:32 +0000 (23:08 +0200)]
gtkquartz: don't free the string returned by get_bundle_path()

It's statically allocated.
(cherry picked from commit c040b03c2e32a773a4d9cf4019050c2f8a5b91ce)

12 years agoquartz: don't return FALSE from GdkDevice::query_state()
Michael Natterer [Sat, 22 Oct 2011 19:32:49 +0000 (21:32 +0200)]
quartz: don't return FALSE from GdkDevice::query_state()

Not finding a child window to return is not a failure.

12 years agoUpdated German translation
Hendrik Knackstedt [Fri, 21 Oct 2011 22:38:28 +0000 (00:38 +0200)]
Updated German translation

12 years agoGtkRange: use the right widget for coordinate translation
Cosimo Cecchi [Fri, 21 Oct 2011 20:30:34 +0000 (16:30 -0400)]
GtkRange: use the right widget for coordinate translation

GtkRange needs to check if its allocation intersects with the resize
grip allocation (trimming its own allocation if it does).
In order to do that, it needs to translate its allocation into window
coordinates, and before that, find the window to whose the allocation
is relative; code goes all the way finding the right parent widget, but
then doesn't actually use it when translating the coordinates, leading
to using the wrong rectangles for the intersection check.

https://bugzilla.gnome.org/show_bug.cgi?id=662308

12 years agoFix document generation in out of tree builds
Sjoerd Simons [Fri, 21 Oct 2011 13:55:50 +0000 (15:55 +0200)]
Fix document generation in out of tree builds

https://bugzilla.gnome.org/show_bug.cgi?id=661032

12 years agoFix natural/min widht calculations for empty grids
Matthias Clasen [Fri, 21 Oct 2011 20:23:03 +0000 (16:23 -0400)]
Fix natural/min widht calculations for empty grids

https://bugzilla.gnome.org/show_bug.cgi?id=660139

12 years agoAdd GtkSwitch to widget-factory
Matthias Clasen [Fri, 21 Oct 2011 02:15:41 +0000 (22:15 -0400)]
Add GtkSwitch to widget-factory

12 years agoDistribute tests/css/parser/selector.errors
Matthias Clasen [Fri, 21 Oct 2011 01:51:21 +0000 (21:51 -0400)]
Distribute tests/css/parser/selector.errors

https://bugzilla.gnome.org/show_bug.cgi?id=662024

12 years agoDon't refer to --sync anymore
Matthias Clasen [Fri, 21 Oct 2011 01:48:54 +0000 (21:48 -0400)]
Don't refer to --sync anymore

That commandline option got replaced with an environment variable,
so mention that in the warning text.

https://bugzilla.gnome.org/show_bug.cgi?id=662132

12 years agoDND: Work better with XI2
Matthias Clasen [Fri, 21 Oct 2011 00:59:50 +0000 (20:59 -0400)]
DND: Work better with XI2

It turns out that simply using XIGrabKeycode instead of XGrabKey
makes the DND keyboard support mostly work (there seem to be some
minor issues with modifiers). This means we no longer grab the
keyboard actively during DND, which in turn makes Alt-Tab and
other window manager shortcuts work again during DND.

At the same time, bring the DND key handling code into the
multi-backend work, by checking for X11 and XI2 at runtime,
in addition to compile time.

12 years agotreemodel: document gtk_tree_path_free() as allow-none
Cosimo Cecchi [Wed, 19 Oct 2011 18:47:50 +0000 (14:47 -0400)]
treemodel: document gtk_tree_path_free() as allow-none

https://bugzilla.gnome.org/show_bug.cgi?id=662230

12 years agogtk: GtkToolbar returns invalid paths when adding tool items
Stef Walter [Wed, 19 Oct 2011 11:17:18 +0000 (13:17 +0200)]
gtk: GtkToolbar returns invalid paths when adding tool items

 * Calculate siblings_path on demand, so that it's always correct
   and doesn't get out of sync with the toolbar state.

https://bugzilla.gnome.org/show_bug.cgi?id=662177

12 years agotooltip: add support for opacity and rounded corners for tooltips
Cosimo Cecchi [Thu, 6 Oct 2011 21:06:57 +0000 (17:06 -0400)]
tooltip: add support for opacity and rounded corners for tooltips

Rounded corners now will always work, using XShape in case we're not
running a composite manager.
Also, setting an RGBA visual (if available) on the tooltip toplevel
enables them to be transparent if the theme specifies so.

https://bugzilla.gnome.org/show_bug.cgi?id=599617

12 years agoGMountOperation on OpenBSD: remove the need for kvm(3)
Antoine Jacoutot [Sat, 15 Oct 2011 09:27:47 +0000 (11:27 +0200)]
GMountOperation on OpenBSD: remove the need for kvm(3)

kvm(3) is considered a deprecated interface, so make
GMountOperation::show-processes use the recommended sysctl(3) interface
instead. This also removes the need to link with libkvm.

https://bugzilla.gnome.org/show_bug.cgi?id=661835

12 years agoFixed some strings in Brazilian Portuguese translation
André Gondim [Mon, 17 Oct 2011 12:03:01 +0000 (10:03 -0200)]
Fixed some strings in Brazilian Portuguese translation

12 years agoUpdate VS property sheets
Chun-wei Fan [Mon, 17 Oct 2011 07:34:24 +0000 (15:34 +0800)]
Update VS property sheets

Bid farewell to G_DISABLE_DEPRECATED and the build errors it causes as
warnings are now used to deter people from using deprecated GLib items
in a more subtle manner

12 years agoUpdated Japanese translation
Jiro Matsuzawa [Sun, 16 Oct 2011 16:25:41 +0000 (01:25 +0900)]
Updated Japanese translation

12 years agoUpdated British English translation
Bruce Cowan [Sun, 16 Oct 2011 13:45:05 +0000 (14:45 +0100)]
Updated British English translation

12 years agoBug 658772: Directory paths for resource directories are hard coded.
John Ralls [Fri, 14 Oct 2011 23:39:37 +0000 (16:39 -0700)]
Bug 658772: Directory paths for resource directories are hard coded.

Corrected formatting to match coding standards; introduced local statics
to prevent leaking returned strings.

12 years agogtk: fix syntax of an annotation
Torsten Schönfeld [Sat, 15 Oct 2011 17:37:43 +0000 (19:37 +0200)]
gtk: fix syntax of an annotation

12 years agoUse G_VALUE_INIT
Javier Jardón [Fri, 30 Sep 2011 15:31:04 +0000 (16:31 +0100)]
Use G_VALUE_INIT

Instead of an explicit { 0, } when declaring the variable.

12 years agogdk: s/gdk_cursor_unref/g_object_unref/
Michael Natterer [Sat, 15 Oct 2011 12:35:28 +0000 (14:35 +0200)]
gdk: s/gdk_cursor_unref/g_object_unref/

12 years agoFixed referenced images for GTK_STOCK_INDENT
Sebastian Pölsterl [Fri, 14 Oct 2011 15:40:01 +0000 (17:40 +0200)]
Fixed referenced images for GTK_STOCK_INDENT

https://bugzilla.gnome.org/show_bug.cgi?id=661780

12 years agoAdd deprecation annotations for GtkHandleBox and GtkTable
Matthias Clasen [Fri, 14 Oct 2011 04:54:01 +0000 (00:54 -0400)]
Add deprecation annotations for GtkHandleBox and GtkTable

12 years agobuild: Strip out -DG_DISABLE_DEPRECATED
Colin Walters [Wed, 12 Oct 2011 15:54:56 +0000 (11:54 -0400)]
build: Strip out -DG_DISABLE_DEPRECATED

It's unnecessary pain now that they're warnings.

12 years agoBug 660730: Win32: Only use _fstat32 if available
Chun-wei Fan [Wed, 12 Oct 2011 03:39:06 +0000 (11:39 +0800)]
Bug 660730: Win32: Only use _fstat32 if available

_fstat32 is only introduced with msvcrt80.dll (i.e. Visual C++ 2005), in
which using this function will break compilation with MinGW, which links
against msvcrt.dll.  The msvcrt.lib in the Windows DDK which links to
a later incarnation of the msvcrt.dll in later Windows systems may have
this symbol defined, but that needs to be checked upon to be sure.

Thanks to Dieter Verfaillie for pointing out this problem.

12 years agoAdd deprecation annotations for deprecated functions
Matthias Clasen [Tue, 11 Oct 2011 19:03:46 +0000 (15:03 -0400)]
Add deprecation annotations for deprecated functions

We define our own GDK_DEPRECATED[_FOR] macros for this
and allow it to be turned off by defining the
GDK_DISABLE_DEPRECATION_WARNINGS macro.

12 years agoInclude gdk.h in gail.c
Matthias Clasen [Tue, 11 Oct 2011 19:01:41 +0000 (15:01 -0400)]
Include gdk.h in gail.c

This is needed so that we pick up deprecation macros here in the
future.

12 years agoAdd a new symbol
Matthias Clasen [Mon, 10 Oct 2011 15:02:33 +0000 (11:02 -0400)]
Add a new symbol

gtk_widget_get_modifier_mask was added recently.

12 years agoFix void function n gtkfontbutton.c so it does not try to return a value.
Brian Cameron [Mon, 10 Oct 2011 19:13:35 +0000 (14:13 -0500)]
Fix void function n gtkfontbutton.c so it does not try to return a value.
 Fixes bug #661089.

12 years agoBug 658767 - Drag and Drop NSEvent capture is racy
John Ralls [Sun, 25 Sep 2011 01:19:56 +0000 (18:19 -0700)]
Bug 658767 - Drag and Drop NSEvent capture is racy

Create a synthetic NSMouseLeftDown to store in the GtkQuartzDragSourceInfo
rather than relying on the NSWindow's latest event being the right one (or the
right kind).

12 years agoBug 657770 - Write to released memory in gtkdnd-quartz.c
Kristian Rietveld [Sun, 25 Sep 2011 00:32:21 +0000 (17:32 -0700)]
Bug 657770 - Write to released memory in gtkdnd-quartz.c

Clear the Drag paste board just before the info->context is released.
This way the GtkDragSourceOwner is released just before the drag context
is and thus can pasteboard:provideDataForType: not accidentally access
an already released drag context

12 years agoQuartz, Bug 655057: Eliminate Compiler Warning
Alex Corrado [Sun, 9 Oct 2011 16:52:14 +0000 (09:52 -0700)]
Quartz, Bug 655057: Eliminate Compiler Warning

When compiled with older SDKs, the original change for this bug caused a
compiler warning about NSWindow not being able to handle a setStyleMask
message. This tricks the compiler into thinking that it can.

12 years agoAdd G_GNUC_DEPRECATED annotations in GDK
Matthias Clasen [Sun, 9 Oct 2011 13:55:08 +0000 (09:55 -0400)]
Add G_GNUC_DEPRECATED annotations in GDK

12 years agoBug 658772: Directory paths for resource directories are hard coded.
John Ralls [Sun, 11 Sep 2011 23:45:06 +0000 (16:45 -0700)]
Bug 658772: Directory paths for resource directories are hard coded.

Provide dynamic path discovery functions as are provided for win32.

12 years agoquartz: also update window position on windowDidResize
Kristian Rietveld [Sat, 8 Oct 2011 09:47:11 +0000 (11:47 +0200)]
quartz: also update window position on windowDidResize

When a window goes fullscreen, the resize also changes its size.

12 years agoUpdated Dutch translation by Wouter Bolsterlee
Wouter Bolsterlee [Fri, 7 Oct 2011 22:40:12 +0000 (00:40 +0200)]
Updated Dutch translation by Wouter Bolsterlee

12 years agoUpdated Dutch translation by Rachid
Rachid [Fri, 7 Oct 2011 22:39:18 +0000 (00:39 +0200)]
Updated Dutch translation by Rachid

12 years agogtk: remove the private GTK_DEFAULT_ACCEL_MOD_MASK define
Michael Natterer [Fri, 7 Oct 2011 18:58:58 +0000 (20:58 +0200)]
gtk: remove the private GTK_DEFAULT_ACCEL_MOD_MASK define

and finish the port to using the new modifier abstraction API.
This commit has some evilness, it uses the default display for
the lack of a widget context, and the change to gtkstock.c
is very ugly, but I can't think of anything better given
GtkStockItem needs an accel mask instead of a proper accel
string.

12 years agogdk: fix gdk_keyval_to_lower/upper() for Quartz, Win32 and Broadway
Michael Natterer [Fri, 7 Oct 2011 14:10:19 +0000 (16:10 +0200)]
gdk: fix gdk_keyval_to_lower/upper() for Quartz, Win32 and Broadway

In 2.x, the !HAVE_XCONVERTCASE fallback of keyval_convert_case() was
implicitly used as implementation for all !X11 backends.

In 3.x, when this function was virtualized in GdkDisplayManager,
this fallback was moved to the X11 backend and the other backends
"equipped" with /* FIXME implement */ implementations of
keyval_convert_case() which don't convert anything.

Move the fallback code back to gdk/ as default implementation
of GdkDisplayManager::keyval_convert_case() and remove its
implementations is all backends but X11. Also remove the
implementation in Wayland which was a plain copy of what
is now the default implementation.
(cherry picked from commit f46c1b76d839a03bdc8080bc130fabd1068f2557)

12 years agogtk: use the new modifier API instead of GTK_DEFAULT_ACCEL_MOD_MASK
Michael Natterer [Fri, 7 Oct 2011 07:29:43 +0000 (09:29 +0200)]
gtk: use the new modifier API instead of GTK_DEFAULT_ACCEL_MOD_MASK

in gtkimcontextsimple and gtktreeview, which are the easy places,
but also true bugfixes, because on OSX we can't use the virtual META
here, we needs MOD2 as delivered in key events.

12 years agoUpdated Galician translations
Fran Dieguez [Thu, 6 Oct 2011 23:25:04 +0000 (01:25 +0200)]
Updated Galician translations

12 years agoGtkComboBox: Add the format-entry-text signal.
Tristan Van Berkom [Thu, 6 Oct 2011 09:14:38 +0000 (11:14 +0200)]
GtkComboBox: Add the format-entry-text signal.

This provides more control over how the selected item is shown in
the entry.
Bug #631167

12 years agoGtkTreeViewColumn: dispose: Remove from the treeview
Kjell Ahlstedt [Wed, 28 Sep 2011 11:02:20 +0000 (13:02 +0200)]
GtkTreeViewColumn: dispose: Remove from the treeview

This handles the case (with gtkmm) that the GtkTreeViewColumn is
destroyed before the GtkTreeView is destroyed.
Bug #659672

12 years agoBug 628936: Minimal change to pass introspection.
John Ralls [Sun, 21 Nov 2010 21:06:57 +0000 (13:06 -0800)]
Bug 628936: Minimal change to pass introspection.

gobject_introspection's G-ir-scanner doesn't like the -xobjective-c
option needed to compile ObjectiveC features in quartz implementations
of certain gtk functions. This rearranges the compiler flag environment
variables in Makefile so that G-ir-scanner doesn't see them.

12 years agoFixed gtk_cell_area_stop_editing() to properly notify when editing is not canceled
Tristan Van Berkom [Mon, 3 Oct 2011 23:32:00 +0000 (19:32 -0400)]
Fixed gtk_cell_area_stop_editing() to properly notify when editing is not canceled

This fixes bug 653289, in gtk+-2 gtk_tree_view_stop_editing() was previously
responsible for notifying with gtk_cell_editable_editing_done(), this detail
was missed when porting the cell renderer management to GtkCellArea.

12 years agoreftests: add a reftest for the CSS :only-child selector
Cosimo Cecchi [Tue, 4 Oct 2011 15:46:25 +0000 (11:46 -0400)]
reftests: add a reftest for the CSS :only-child selector

https://bugzilla.gnome.org/show_bug.cgi?id=660833

12 years agoall: add an 'only-child' pseudo-class selector
Cosimo Cecchi [Mon, 3 Oct 2011 21:09:09 +0000 (17:09 -0400)]
all: add an 'only-child' pseudo-class selector

Add an :only-child pseudo-class selector, which selects single children
of a parent container.

https://bugzilla.gnome.org/show_bug.cgi?id=660833

12 years agoappchooser: make the "Find applications online" follow search state
Cosimo Cecchi [Mon, 3 Oct 2011 15:50:53 +0000 (11:50 -0400)]
appchooser: make the "Find applications online" follow search state

Set the button as insensitive while searching, so that it can't be
accidentally clicked twice.

https://bugzilla.gnome.org/show_bug.cgi?id=651818

12 years agoappchooser: add a GCancellable to the search operation
Cosimo Cecchi [Mon, 3 Oct 2011 15:47:09 +0000 (11:47 -0400)]
appchooser: add a GCancellable to the search operation

So that we can cancel it when the dialog is destroyed.

https://bugzilla.gnome.org/show_bug.cgi?id=651818

12 years agoUpdated Spanish translation
Jorge González [Mon, 3 Oct 2011 21:21:40 +0000 (23:21 +0200)]
Updated Spanish translation

12 years agoFix commit 61d6c73f186a08cd55852c0a44a1704a6fbbbc39
Kristian Rietveld [Mon, 3 Oct 2011 20:57:12 +0000 (22:57 +0200)]
Fix commit 61d6c73f186a08cd55852c0a44a1704a6fbbbc39

Where the commit suggested "somebody should review it [the test]",
it would have been useful to test this particular commit instead...

12 years agoBug 659022 - gtk_tree_model_filter_clear_cache_helper: assertion failed
Kristian Rietveld [Mon, 3 Oct 2011 20:38:42 +0000 (22:38 +0200)]
Bug 659022 - gtk_tree_model_filter_clear_cache_helper: assertion failed

This bug is resolved by fixing two things in
gtk_tree_model_filter_row_deleted():

  (1) It is possible for an elt to have elt->visible_siter == NULL, when
  it is deleted.  Only call g_sequence_remove() if this pointer is
  non-NULL.

  (2) For the case len (level->seq) > 1, free the elt->children level
  if non-NULL.  Failing to do this means the level will stick around.
  If this child level was not referenced, it will still have a zero
  ref count on its parent which cannot be removed!

For both bugs unit tests have been added in the preceding commit.

12 years agoAdd two more unit tests for bug 659022
Kristian Rietveld [Mon, 3 Oct 2011 20:37:49 +0000 (22:37 +0200)]
Add two more unit tests for bug 659022

Both pointing out problems in gtk_tree_model_filter_row_deleted().

12 years agoFix first part of bug 659022
Kristian Rietveld [Mon, 3 Oct 2011 19:35:21 +0000 (21:35 +0200)]
Fix first part of bug 659022

Push creation of path into if clause, the path cannot be created anyway
(and would be meaningless otherwise) if the parent is not visible.

12 years agoAdd unit test for part 1 of bug 659022
Kristian Rietveld [Mon, 3 Oct 2011 19:33:30 +0000 (21:33 +0200)]
Add unit test for part 1 of bug 659022

This tests a case where gtk_tree_model_filter_row_changed() is run
for a node with visible_siter != NULL, however its parent has
visible_iter == NULL. For this case, no valid paths can be generated
as the node cannot be accessed (because the parent is invisible).

12 years agobox: add some missing includes
Cosimo Cecchi [Mon, 3 Oct 2011 20:54:09 +0000 (16:54 -0400)]
box: add some missing includes

12 years agoiconcache: Fix gcc warning
Benjamin Otte [Mon, 3 Oct 2011 16:03:57 +0000 (18:03 +0200)]
iconcache: Fix gcc warning

12 years agofilesystemmodel: Add a missing check
Benjamin Otte [Sun, 2 Oct 2011 03:27:41 +0000 (05:27 +0200)]
filesystemmodel: Add a missing check

Paths with depth > 1 should return FALSE instead of pretending the depth
is 1.

12 years agoBug 660730: Use GStatBuf for portability
Chun-wei Fan [Mon, 3 Oct 2011 15:25:33 +0000 (23:25 +0800)]
Bug 660730: Use GStatBuf for portability

Thanks to Kean Johnston for pointing this out.

There are a few places in GTK that use "struct stat",
and then g_stat(), rather than using GStatBuf.This breaks things on
Windows. Since the size of struct stat can vary depending on other
flags specified, this has the potential to cause overwrites and is
trivial to fix.

Based on patch submitted by Kean Johnston

12 years agogdk: Plug leak in gdk_display_pointer_is_grabbed
Adel Gadllah [Sun, 2 Oct 2011 07:52:16 +0000 (09:52 +0200)]
gdk: Plug leak in gdk_display_pointer_is_grabbed

gdk_x11_device_manager_core_list_devices returns a new allocated
list, which has to be freed.

valgrind output:

==18686== 160,176 (80,088 direct, 80,088 indirect) bytes in 3,337 blocks are definitely lost in loss record 25,347 of 25,378
==18686==    at 0x4C256DD: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==18686==    by 0x6CD7752: g_malloc (in /lib64/libglib-2.0.so.0.3000.0)
==18686==    by 0x6CEE2B6: g_slice_alloc (in /lib64/libglib-2.0.so.0.3000.0)
==18686==    by 0x6CCB37D: g_list_prepend (in /lib64/libglib-2.0.so.0.3000.0)
==18686==    by 0x654CADA: gdk_x11_device_manager_core_list_devices (gdkdevicemanager-core-x11.c:836)
==18686==    by 0x6531489: gdk_display_pointer_is_grabbed (gdkdisplay.c:1270)
==18686==    by 0x5162E1E: filter_func (ui.c:140)
==18686==    by 0x6558B50: gdk_event_apply_filters (gdkeventsource.c:83)
==18686==    by 0x6558CB3: _gdk_x11_display_queue_events (gdkeventsource.c:197)
==18686==    by 0x6530680: gdk_display_get_event (gdkdisplay.c:311)
==18686==    by 0x65589F1: gdk_event_source_dispatch (gdkeventsource.c:356)
==18686==    by 0x6CD0A0E: g_main_context_dispatch (in /lib64/libglib-2.0.so.0.3000.0)

https://bugzilla.gnome.org/show_bug.cgi?id=660676

12 years agoUpdated Esperanto translation
Tiffany Antopolski [Mon, 3 Oct 2011 10:48:57 +0000 (12:48 +0200)]
Updated Esperanto translation

12 years ago__SIZEOF_INT__ and __SIZEOF_POINTER__ may not be defined.
Antoine Jacoutot [Mon, 3 Oct 2011 07:11:50 +0000 (09:11 +0200)]
__SIZEOF_INT__ and __SIZEOF_POINTER__ may not be defined.

Protect __SIZEOF_INT__ and __SIZEOF_POINTER__ with an ifdef to prevent
warnings about them not being defined when including gtktextattributes.h.

https://bugzilla.gnome.org/show_bug.cgi?id=660619

12 years agogtkbox: Move private functions to private header
Javier Jardón [Sun, 2 Oct 2011 20:21:49 +0000 (21:21 +0100)]
gtkbox: Move private functions to private header

12 years agoAdd 'Since 3.4' indices to the docs
Matthias Clasen [Sun, 2 Oct 2011 14:20:48 +0000 (10:20 -0400)]
Add 'Since 3.4' indices to the docs

12 years agoAdd new api to gdk.symbols
Matthias Clasen [Sun, 2 Oct 2011 14:19:15 +0000 (10:19 -0400)]
Add new api to gdk.symbols

This keeps 'make check' happy.
And if 'make check' is happy, so am I.

12 years agogdkcreen-x11: Use GdkRGBA instead GdkColor
Javier Jardón [Mon, 13 Jun 2011 00:54:02 +0000 (01:54 +0100)]
gdkcreen-x11: Use GdkRGBA instead GdkColor

12 years agoUpdated Vietnamese translation
Nguyễn Thái Ngọc Duy [Sat, 1 Oct 2011 10:10:12 +0000 (20:10 +1000)]
Updated Vietnamese translation

12 years agopo/vi: import from Damned Lies
Nguyễn Thái Ngọc Duy [Sat, 1 Oct 2011 09:55:34 +0000 (19:55 +1000)]
po/vi: import from Damned Lies

12 years agouimanager: Deprecate add-tearoffs property
Benjamin Otte [Sat, 1 Oct 2011 03:58:27 +0000 (05:58 +0200)]
uimanager: Deprecate add-tearoffs property

Don't use it. Please.

12 years agomenu: Deprecate GtkTearoffMenuItem
Benjamin Otte [Sat, 1 Oct 2011 03:49:48 +0000 (05:49 +0200)]
menu: Deprecate GtkTearoffMenuItem

12 years agotestgtk: Remove tearoff menus from tests
Benjamin Otte [Sat, 1 Oct 2011 03:49:30 +0000 (05:49 +0200)]
testgtk: Remove tearoff menus from tests

12 years agogtk-demo: Don't use tearoff menus
Benjamin Otte [Sat, 1 Oct 2011 03:49:00 +0000 (05:49 +0200)]
gtk-demo: Don't use tearoff menus

Demoing bad things, sheesh.

12 years agogtkwindow: Clean up gtk_window_map()
Rui Matos [Sat, 1 Oct 2011 00:18:52 +0000 (01:18 +0100)]
gtkwindow: Clean up gtk_window_map()

Remove unneeded variable and delete trailing whitespace.

12 years agodocs: Delete duplicate line in gdkwindow.c
Rui Matos [Sat, 1 Oct 2011 00:15:45 +0000 (01:15 +0100)]
docs: Delete duplicate line in gdkwindow.c

12 years agohandlebox: Deprecate
Benjamin Otte [Fri, 30 Sep 2011 20:31:56 +0000 (22:31 +0200)]
handlebox: Deprecate

12 years agotestgtk: Remove handlebox example
Benjamin Otte [Fri, 30 Sep 2011 20:10:18 +0000 (22:10 +0200)]
testgtk: Remove handlebox example

Rather than keeping it and enabling deprecated code in testgtk I decided
to delete the example. The warnings from testgtk about deprecated code
are too useful IMO.
Should we need it again, we can make it into its own test, or add a
testgtkdeprecated.

12 years agotests: Remove handleboxes from testmerge example
Benjamin Otte [Fri, 30 Sep 2011 20:10:04 +0000 (22:10 +0200)]
tests: Remove handleboxes from testmerge example

12 years agoa11y: Don't include gtkx when not needed
Benjamin Otte [Fri, 30 Sep 2011 19:40:39 +0000 (21:40 +0200)]
a11y: Don't include gtkx when not needed

12 years agoa11y: Remove useless check
Benjamin Otte [Fri, 30 Sep 2011 19:39:08 +0000 (21:39 +0200)]
a11y: Remove useless check

12 years agoUpdated Hebrew translation.
Yaron Shahrabani [Fri, 30 Sep 2011 10:27:54 +0000 (13:27 +0300)]
Updated Hebrew translation.