gtk4.git
16 years agoCorrect the initial event mask of the root window
Matthias Clasen [Sun, 6 Sep 2009 06:42:39 +0000 (02:42 -0400)]
Correct the initial event mask of the root window

This fixes bug 588455 where the panel stopped responding to Alt-F2 when
setting a transparent background.

16 years agoFix compiler warnings
Kristian Rietveld [Sat, 5 Sep 2009 14:59:17 +0000 (16:59 +0200)]
Fix compiler warnings

16 years agoFix memleak in filter model unit test
Kristian Rietveld [Sat, 5 Sep 2009 14:58:12 +0000 (16:58 +0200)]
Fix memleak in filter model unit test

(There is one larger memleak left here, but this is a side effect of the
test case for bug 346800).

16 years agoList work left to do in the filter model unit test
Kristian Rietveld [Sat, 5 Sep 2009 14:47:35 +0000 (16:47 +0200)]
List work left to do in the filter model unit test

16 years agoEmit row-has-child-toggled when a first node becomes visible in a level
Kristian Rietveld [Sat, 5 Sep 2009 14:41:09 +0000 (16:41 +0200)]
Emit row-has-child-toggled when a first node becomes visible in a level

If we have a level with zero visible nodes and the first node becomes
visible in that level, then parent has just become a "real" parent node.
In such a case we need to emit row-has-child-toggled.  This only applies
to non-root levels that have a parent.  This problem was also found when
writing the unit test, the respective cases in the unit test have been
corrected.

This fixes bugs:
 Bug 372010 - Filtering not working properly
 Bug 525965 - Filtered and sorted GtkTreeView is missing rows

16 years agoWhen the filter removes a node, it needs to remove its children too
Kristian Rietveld [Sat, 5 Sep 2009 14:12:48 +0000 (16:12 +0200)]
When the filter removes a node, it needs to remove its children too

In gtk_tree_model_filter_remove_node(), we check if the given elt has
children.  If so, this level will be freed.  This action is recursive.

16 years agoHave the unit test check if the filter model emits the right signals
Kristian Rietveld [Sat, 5 Sep 2009 13:57:25 +0000 (15:57 +0200)]
Have the unit test check if the filter model emits the right signals

16 years agoProperly handle cases where an ancestor node is invisible
Kristian Rietveld [Fri, 4 Sep 2009 07:28:21 +0000 (09:28 +0200)]
Properly handle cases where an ancestor node is invisible

16 years agoExtend filter model unit test with row-has-child-toggled testers
Kristian Rietveld [Wed, 2 Sep 2009 10:47:53 +0000 (12:47 +0200)]
Extend filter model unit test with row-has-child-toggled testers

16 years agoCorrect handling of row-has-child-toggled signal
Kristian Rietveld [Wed, 2 Sep 2009 10:45:41 +0000 (12:45 +0200)]
Correct handling of row-has-child-toggled signal

When a filter function is used that determines the visibility property
of a node using whether or it it has children, the state of this node
very likely changes when the filter model receives a
row-has-child-toggled signal.  Therefore, we need to handle state
changes in the row-has-child-toggled handler.

16 years agoRemove unused variable
Kristian Rietveld [Tue, 1 Sep 2009 16:06:56 +0000 (18:06 +0200)]
Remove unused variable

16 years agoMore virtual root unit tests
Kristian Rietveld [Tue, 1 Sep 2009 16:05:46 +0000 (18:05 +0200)]
More virtual root unit tests

16 years agoAdd a rather specially crafted test case
Kristian Rietveld [Tue, 1 Sep 2009 15:51:50 +0000 (17:51 +0200)]
Add a rather specially crafted test case

16 years agoMarking nodes invisible after pulling in the root level is wrong
Kristian Rietveld [Tue, 1 Sep 2009 15:51:19 +0000 (17:51 +0200)]
Marking nodes invisible after pulling in the root level is wrong

16 years agoDo not emit has-child-toggled when a root level node is removed
Kristian Rietveld [Tue, 1 Sep 2009 15:49:02 +0000 (17:49 +0200)]
Do not emit has-child-toggled when a root level node is removed

16 years agoCorrect unfiltered test cases
Kristian Rietveld [Tue, 1 Sep 2009 14:45:39 +0000 (16:45 +0200)]
Correct unfiltered test cases

16 years agoCheck for successfull iterator retrieval
Kristian Rietveld [Tue, 1 Sep 2009 14:37:57 +0000 (16:37 +0200)]
Check for successfull iterator retrieval

16 years agoRefilter is required after setting visible column
Kristian Rietveld [Tue, 1 Sep 2009 14:19:30 +0000 (16:19 +0200)]
Refilter is required after setting visible column

16 years agoCorrect /FilterModel/empty/vroot-show-nodes test case
Kristian Rietveld [Tue, 1 Sep 2009 14:15:08 +0000 (16:15 +0200)]
Correct /FilterModel/empty/vroot-show-nodes test case

16 years agoOnly emit row-inserted if all ancestors are visible
Kristian Rietveld [Tue, 1 Sep 2009 14:10:05 +0000 (16:10 +0200)]
Only emit row-inserted if all ancestors are visible

16 years agoGeneralize fixture setup code
Kristian Rietveld [Tue, 1 Sep 2009 12:50:32 +0000 (14:50 +0200)]
Generalize fixture setup code

16 years agoMake a start with virtual root unit tests
Kristian Rietveld [Tue, 1 Sep 2009 12:46:11 +0000 (14:46 +0200)]
Make a start with virtual root unit tests

16 years agoAdd test cases for bugs 464173 and 529287
Kristian Rietveld [Tue, 1 Sep 2009 11:54:35 +0000 (13:54 +0200)]
Add test cases for bugs 464173 and 529287

16 years agoAdd unit test for bug 540201
Kristian Rietveld [Tue, 1 Sep 2009 09:14:40 +0000 (11:14 +0200)]
Add unit test for bug 540201

16 years agoAdd a helper for inserting paths in the test tree store
Kristian Rietveld [Tue, 1 Sep 2009 07:39:11 +0000 (09:39 +0200)]
Add a helper for inserting paths in the test tree store

16 years agoStart a testsuite for GtkTreeModelFilter
Kristian Rietveld [Tue, 1 Sep 2009 07:17:02 +0000 (09:17 +0200)]
Start a testsuite for GtkTreeModelFilter

16 years agoFix fallout from automake version bump
Matthias Clasen [Sat, 5 Sep 2009 05:46:39 +0000 (01:46 -0400)]
Fix fallout from automake version bump

16 years agoBump version
Matthias Clasen [Sat, 5 Sep 2009 04:36:49 +0000 (00:36 -0400)]
Bump version

16 years ago2.17.11
Matthias Clasen [Sat, 5 Sep 2009 04:35:08 +0000 (00:35 -0400)]
2.17.11

16 years agoFix a typo
Matthias Clasen [Sat, 5 Sep 2009 03:27:37 +0000 (23:27 -0400)]
Fix a typo

16 years agoFix the build
Matthias Clasen [Sat, 5 Sep 2009 03:00:27 +0000 (23:00 -0400)]
Fix the build

16 years agoDocumentation additions
Matthias Clasen [Sat, 5 Sep 2009 02:53:03 +0000 (22:53 -0400)]
Documentation additions

16 years agoUpdates
Matthias Clasen [Sat, 5 Sep 2009 02:47:59 +0000 (22:47 -0400)]
Updates

16 years agoMake gtk_combo_box_entry_set_text_column more forgiving
Matthias Clasen [Sat, 5 Sep 2009 02:20:28 +0000 (22:20 -0400)]
Make gtk_combo_box_entry_set_text_column more forgiving

The assertions in there were unnecessarily restrictive. This has
been reported in bug 564160.

16 years agoDon't forget to set the client window on the slave
Matthias Clasen [Sat, 5 Sep 2009 00:34:09 +0000 (20:34 -0400)]
Don't forget to set the client window on the slave

16 years agoMake window manager tracking work better
Matthias Clasen [Sat, 5 Sep 2009 00:22:43 +0000 (20:22 -0400)]
Make window manager tracking work better

We were getting the new wm name with a 15 second delay, due to
some race. Reported in bug 593644.

16 years agoAvoid a segfault if license is NULL
Matthias Clasen [Fri, 4 Sep 2009 22:22:30 +0000 (18:22 -0400)]
Avoid a segfault if license is NULL

The license field was introduced later on, so some out-of-tree loaders
may not set it. And Solaris printf() doesn't take NULL for a string...
Bug 594178.

16 years agoRework the way IM contexts are set
Matthias Clasen [Fri, 4 Sep 2009 22:09:44 +0000 (18:09 -0400)]
Rework the way IM contexts are set

Add a function to obtain the effective context id, and reset the slave
only when the effective context id is different from the current context
id, when setting a client window and on focus in. This might fix
bug 593868 and bug 567124.

16 years agoTry again to allow newer automake
Matthias Clasen [Fri, 4 Sep 2009 19:13:09 +0000 (15:13 -0400)]
Try again to allow newer automake

16 years agoRevert "Add AM_PROG_CC_C_O"
Matthias Clasen [Fri, 4 Sep 2009 16:05:10 +0000 (12:05 -0400)]
Revert "Add AM_PROG_CC_C_O"

This reverts commit 1606b822321fb4cd370a2cd2e117481e125722ee.

16 years agoRevert "Accept automake 1.10 and 1.11"
Matthias Clasen [Fri, 4 Sep 2009 16:04:51 +0000 (12:04 -0400)]
Revert "Accept automake 1.10 and 1.11"

This reverts commit 76dab7b3b745c285f814311b071fefc8244d5245.

16 years agoDeprecate GDK_WINDOW_OBJECT and GdkWindowObject
Alexander Larsson [Fri, 4 Sep 2009 15:53:27 +0000 (17:53 +0200)]
Deprecate GDK_WINDOW_OBJECT and GdkWindowObject

16 years agoAdd function accessors for x11 window/pixmap impl
Alexander Larsson [Fri, 4 Sep 2009 15:52:46 +0000 (17:52 +0200)]
Add function accessors for x11 window/pixmap impl

16 years agoMake public GDK_WINDOW_TYPE and GDK_WINDOW_DESTROYED use functions calls
Alexander Larsson [Fri, 4 Sep 2009 15:50:42 +0000 (17:50 +0200)]
Make public GDK_WINDOW_TYPE and GDK_WINDOW_DESTROYED use functions calls

16 years agoAdd gdk_window_is_destroyed
Alexander Larsson [Fri, 4 Sep 2009 15:49:57 +0000 (17:49 +0200)]
Add gdk_window_is_destroyed

16 years agoMove _gdk_window_event_parent_of to gdkinternals.h
Alexander Larsson [Fri, 4 Sep 2009 15:48:49 +0000 (17:48 +0200)]
Move _gdk_window_event_parent_of to gdkinternals.h

gdkprivate.h is in installed file and this function does not
need to be exported.

16 years agoAccept automake 1.10 and 1.11
Matthias Clasen [Fri, 4 Sep 2009 15:52:55 +0000 (11:52 -0400)]
Accept automake 1.10 and 1.11

This was tested to work in bug 588788.

16 years agoAdd AM_PROG_CC_C_O
Matthias Clasen [Fri, 4 Sep 2009 15:51:50 +0000 (11:51 -0400)]
Add AM_PROG_CC_C_O

16 years agoProtect gtk_text_view_scroll_mark_onscreen() from using invalid marks
Kristian Rietveld [Fri, 4 Sep 2009 15:40:12 +0000 (17:40 +0200)]
Protect gtk_text_view_scroll_mark_onscreen() from using invalid marks

Suggested by Paolo Borelli.

16 years agoRecalculate heights when a (new) separator func is set on GtkTreeView
Kristian Rietveld [Fri, 4 Sep 2009 14:56:05 +0000 (16:56 +0200)]
Recalculate heights when a (new) separator func is set on GtkTreeView

16 years agoMove mark check to gtk_text_view_scroll_to_mark()
Kristian Rietveld [Fri, 4 Sep 2009 14:38:04 +0000 (16:38 +0200)]
Move mark check to gtk_text_view_scroll_to_mark()

Also use a different means to check that the given mark is in the text
view's buffer.  This will also play nice with anonymous and NULL marks.
Suggested by Paolo Borelli.

16 years agoEnsure clip region valid before using it for emulating graphics exposures
Alexander Larsson [Fri, 4 Sep 2009 14:01:35 +0000 (16:01 +0200)]
Ensure clip region valid before using it for emulating graphics exposures

If there is a drawable clip on it we don't want to use that.

16 years agoEnsure gc clip region valid in _gdk_gc_update_context
Alexander Larsson [Fri, 4 Sep 2009 13:59:22 +0000 (15:59 +0200)]
Ensure gc clip region valid in _gdk_gc_update_context

There might be an old drawable clip on it, if so remove it.

16 years agogtk_widget_set_property(): use accessors instead of setting flags
Michael Natterer [Fri, 4 Sep 2009 13:18:24 +0000 (15:18 +0200)]
gtk_widget_set_property(): use accessors instead of setting flags

Use the newly added accessors to set the "can-focus", "can-default"
and "receives-default" properties.

16 years agogtk_widget_set_property(): use gtk_widget_set_visible()
Michael Natterer [Fri, 4 Sep 2009 12:54:13 +0000 (14:54 +0200)]
gtk_widget_set_property(): use gtk_widget_set_visible()

Use the new API instead of show() and hide().

16 years agoBug 69872 - GTK_WIDGET_SET_FLAGS should be deprecated
Michael Natterer [Fri, 4 Sep 2009 12:50:35 +0000 (14:50 +0200)]
Bug 69872 - GTK_WIDGET_SET_FLAGS should be deprecated

Add gtk_widget_set_receives_default() and
gtk_widget_get_receives_default() as accessors for
GTK_RECEIVES_DEFAULT.

16 years agoImprove unsetting old buffer in gtk_text_view_set_buffer()
Kristian Rietveld [Fri, 4 Sep 2009 11:41:17 +0000 (13:41 +0200)]
Improve unsetting old buffer in gtk_text_view_set_buffer()

When unsetting the old buffer always set the buffer on the layout to
NULL.  More importantly, clear the pending scroll.  (The scroll is
handled in an idle, when not cleared an idle handler might touch the
layout later on, possibly corrupting the BTree).  Unref the buffer after
removing the selection from the clipboard, not before.  Patch merged
from maemo-gtk.

16 years agoFail to scroll if the given mark is not in text view's current buffer
Kristian Rietveld [Fri, 4 Sep 2009 11:34:56 +0000 (13:34 +0200)]
Fail to scroll if the given mark is not in text view's current buffer

In gtk_text_view_queue_scroll() we need to verify if the given mark
exists in the text view's current buffer.  When not done, this can
result in corruption of the BTree data structure.  Patch merged from
maemo-gtk.

16 years agoUpdate IM spot location before running _validate_onscreen()
Kristian Rietveld [Fri, 4 Sep 2009 11:30:48 +0000 (13:30 +0200)]
Update IM spot location before running _validate_onscreen()

Updating the IM spot location in gtk_text_view_value_changed() might
invalidate the layout, so we need to make sure that we update it before
validating the layout again.  Otherwise, the layout will be invalidated
right after validating it (possibly resulting in a failed
onscreen_validated assertion).  Patch merged from maemo-gtk.

16 years agoMake toggle cell renderer follow insensitive state of widget
Kristian Rietveld [Fri, 4 Sep 2009 11:21:26 +0000 (13:21 +0200)]
Make toggle cell renderer follow insensitive state of widget

Likewise to other cell renderers, GtkCellRendererToggle now properly
follows the insensitive state of the widget requesting rendering.  It
does this by checking the state of the widget pointer, not by using
GtkCellRendererFlags (there exists an insensitive flags).  Later on, I
think we should move to using GtkCellRendererFlags and not the state of
the widget requesting rendering.  Patch merged from maemo-gtk.

16 years agoPropagate insensitive state to cell renderers in GtkCellView
Kristian Rietveld [Fri, 4 Sep 2009 11:11:19 +0000 (13:11 +0200)]
Propagate insensitive state to cell renderers in GtkCellView

If the cell view is insensitive, it needs to properly propagate this
insensitive state to the cell renderers using cell renderer flags.
Merged from maemo-gtk.

16 years agoHave tree view process exposes after adjustments have been updated
Kristian Rietveld [Fri, 4 Sep 2009 09:06:11 +0000 (11:06 +0200)]
Have tree view process exposes after adjustments have been updated

This was removed long ago by the patch in bug 101235.  On a hindsight,
all other scrollable widgets are processing exposes at this point, which
completely makes sense to me, so I am putting it back.

16 years agoCorrect the antiexposure vs implicit paint flush race
Alexander Larsson [Fri, 4 Sep 2009 11:14:25 +0000 (13:14 +0200)]
Correct the antiexposure vs implicit paint flush race

The check for a possible implicit paint flush before queueing an
antiexposure was wrong. An implicit flush doesn't actually NULL
the implicit paint, se we have add a flag to explicitly track if
it is flushed.

16 years agoDon't access region when it might be freed
Alexander Larsson [Fri, 4 Sep 2009 11:10:35 +0000 (13:10 +0200)]
Don't access region when it might be freed

Passing region into _gdk_gc_set_clip_region_internal takes ownership,
so don't use it after that. We can just as well just move the usage
above the call.

16 years agoFix warnings in gdk_window_restack
Alexander Larsson [Fri, 4 Sep 2009 09:08:46 +0000 (11:08 +0200)]
Fix warnings in gdk_window_restack

16 years agoBug 584638 - Build of gtkupdateiconcache without NLS breaks
Christian Dywan [Thu, 3 Sep 2009 09:30:12 +0000 (11:30 +0200)]
Bug 584638 - Build of gtkupdateiconcache without NLS breaks

Wrap textdomain calls in updateiconcache.c in NLS conditionals.

16 years agoFix the return type of gdk_selection_property_get
Matthias Clasen [Thu, 3 Sep 2009 02:50:05 +0000 (22:50 -0400)]
Fix the return type of gdk_selection_property_get

It returns the length of the property, not a boolean. Reported
in bug 593788.

16 years agoAdd a missing include
Matthias Clasen [Thu, 3 Sep 2009 02:14:47 +0000 (22:14 -0400)]
Add a missing include

This was reported in bug 593606.

16 years agoRemove sincos use
Matthias Clasen [Thu, 3 Sep 2009 02:11:28 +0000 (22:11 -0400)]
Remove sincos use

It turns out to be not worth the portability pain. Bug 593877

16 years agoFix a merge conflict
Matthias Clasen [Thu, 3 Sep 2009 01:39:36 +0000 (21:39 -0400)]
Fix a merge conflict

16 years agoBump version
Matthias Clasen [Tue, 1 Sep 2009 13:25:18 +0000 (09:25 -0400)]
Bump version

16 years ago2.17.10
Matthias Clasen [Tue, 1 Sep 2009 13:23:08 +0000 (09:23 -0400)]
2.17.10

16 years agoHungarian translation updated
Gabor Kelemen [Thu, 3 Sep 2009 00:48:08 +0000 (02:48 +0200)]
Hungarian translation updated

16 years agoHungarian translation updated
Gabor Kelemen [Wed, 2 Sep 2009 22:08:59 +0000 (00:08 +0200)]
Hungarian translation updated

16 years agoAdd test for gdk_window_restack
Alexander Larsson [Wed, 2 Sep 2009 21:39:57 +0000 (23:39 +0200)]
Add test for gdk_window_restack

16 years agoAdd gdk_window_restack
Alexander Larsson [Wed, 2 Sep 2009 21:38:55 +0000 (23:38 +0200)]
Add gdk_window_restack

This lets you restack a window above or below a specified sibling.
At least eclipse wants this functionallity.

16 years agoAdd gdk_cairo_reset_clip
Alexander Larsson [Tue, 1 Sep 2009 15:05:50 +0000 (17:05 +0200)]
Add gdk_cairo_reset_clip

This is required to get back the original drawable clip due to
non-native windows on a cairo_t.

16 years agoMove timer removal in dispose
Paolo Borelli [Mon, 3 Aug 2009 19:39:34 +0000 (21:39 +0200)]
Move timer removal in dispose

Move removal of timer function in dispose to prevent potentially running
them on a diposed-but-not-finalized tooltip

16 years agoUpdated Hebrew translation
Yaron Shahrabani [Tue, 1 Sep 2009 13:17:13 +0000 (16:17 +0300)]
Updated Hebrew translation

16 years agoUpdate Norwegian bokmål translation.
Kjartan Maraas [Tue, 1 Sep 2009 08:49:44 +0000 (10:49 +0200)]
Update Norwegian bokmål translation.

16 years agoUpdated French translation
Claude Paroz [Tue, 1 Sep 2009 07:43:36 +0000 (09:43 +0200)]
Updated French translation

16 years agoDocumentation fixes
Matthias Clasen [Tue, 1 Sep 2009 04:53:44 +0000 (00:53 -0400)]
Documentation fixes

16 years agoMore updates
Matthias Clasen [Tue, 1 Sep 2009 04:01:13 +0000 (00:01 -0400)]
More updates

16 years agoFix the cups checks
Matthias Clasen [Tue, 1 Sep 2009 03:59:44 +0000 (23:59 -0400)]
Fix the cups checks

This was reported in bug 593712.

16 years agoMore updates
Matthias Clasen [Tue, 1 Sep 2009 03:02:49 +0000 (23:02 -0400)]
More updates

16 years agoDocument gdk_window_geometry_changed
Matthias Clasen [Mon, 31 Aug 2009 23:51:19 +0000 (19:51 -0400)]
Document gdk_window_geometry_changed

16 years agoHandle geometry change correctly
Matthias Clasen [Mon, 31 Aug 2009 23:41:09 +0000 (19:41 -0400)]
Handle geometry change correctly

This makes sure that the button is properly prelighted when rotating it
underneath the pointer by using keynav on the scale.

16 years agoAdd another offscreen demo
Matthias Clasen [Mon, 31 Aug 2009 23:36:35 +0000 (19:36 -0400)]
Add another offscreen demo

This demo shows rendering multiple times, to achieve a reflection effect.
Inspired by an earlier demo by macslow.

16 years agoMake sure the paper dialog is destroyed with the parent
Paolo Borelli [Mon, 31 Aug 2009 19:56:14 +0000 (21:56 +0200)]
Make sure the paper dialog is destroyed with the parent

16 years agoRemove unneeded call of begin/end user action
Paolo Borelli [Sun, 16 Aug 2009 00:36:15 +0000 (02:36 +0200)]
Remove unneeded call of begin/end user action

delete_interactive already uses them so no need to wrap the call.

16 years agoFix order of cairo clip setting when drawing text
Alexander Larsson [Mon, 31 Aug 2009 14:37:50 +0000 (16:37 +0200)]
Fix order of cairo clip setting when drawing text

We need to set the window clip region before applying the gc clip region,
otherwise we will reset the gc clip region. Fixes bug 593595

16 years agoDescribe how gdk_x11_drawable_get_xid() works wrt CSW in docs
Alexander Larsson [Mon, 31 Aug 2009 14:16:23 +0000 (16:16 +0200)]
Describe how gdk_x11_drawable_get_xid() works wrt CSW in docs

16 years agoUpdated Telugu Translations
krishnababu k [Mon, 31 Aug 2009 14:18:13 +0000 (19:48 +0530)]
Updated Telugu Translations

16 years agoUpdated Telugu Translations
krishnababu k [Mon, 31 Aug 2009 14:17:32 +0000 (19:47 +0530)]
Updated Telugu Translations

16 years agoDon't delay showing windows until viewable in GDK_NATIVE_WINDOWS mode
Alexander Larsson [Fri, 28 Aug 2009 10:10:49 +0000 (12:10 +0200)]
Don't delay showing windows until viewable in GDK_NATIVE_WINDOWS mode

If all windows are native there is no need for delaying the operations
so we just do all shows/hides directly.

16 years agoDon't propagate GDK_NATIVE_WINDOWS to child processes
Alexander Larsson [Mon, 31 Aug 2009 07:38:14 +0000 (09:38 +0200)]
Don't propagate GDK_NATIVE_WINDOWS to child processes

Generally you only need to work around bugs in one specific app, so we
don't want to affect the applications that application will start.
Thus we unset GDK_NATIVE_WINDOWS after reading it.

16 years agoUpdate shape handling
Alexander Larsson [Thu, 27 Aug 2009 16:04:07 +0000 (18:04 +0200)]
Update shape handling

For toplevels, never apply clip as shape, instead apply shape.
This way we don't have to re-set it all the time as the window size
changes. Furthermore, this change fixes unsetting a shape on a
toplevel window which didn't actually unset the shape before.

Additionally we never apply clips as shape if the shape would just
be the same as the regular window size. This means we won't unnecessarily
add a useless shape to most native child windows (and additionally this
helps apps that do weird X stuff that don't expect these shaped windows).

16 years agoDon't update window->shaped from backends
Alexander Larsson [Fri, 28 Aug 2009 10:03:04 +0000 (12:03 +0200)]
Don't update window->shaped from backends

This is properly updated from the main code these days and
should not be touched by the backend code.

16 years agoUse temporary impl_iface variable to make lines shorter
Alexander Larsson [Mon, 31 Aug 2009 08:44:11 +0000 (10:44 +0200)]
Use temporary impl_iface variable to make lines shorter

I.e. we use:
 impl_iface = GDK_WINDOW_IMPL_GET_IFACE (private->impl);
And then use impl_iface instead of the full macro when calling vfuncs.

Also, in some places we avoid getting the iface multiple times.

16 years agoAdd and use gdk_window_is_toplevel when possible
Alexander Larsson [Mon, 31 Aug 2009 08:59:21 +0000 (10:59 +0200)]
Add and use gdk_window_is_toplevel when possible

This makes a lot of code cleaner and easier to understand