]> Pileus Git - ~andy/gtk/log
~andy/gtk
13 years agoFixed GtkTreeModel::row-deleted documentation
Tristan Van Berkom [Wed, 9 Mar 2011 06:06:45 +0000 (15:06 +0900)]
Fixed GtkTreeModel::row-deleted documentation

This documentation was in direct conflict with the documentation
of gtk_tree_model_row_deleted() docs, the function docs are TRUE,
the signal docs were FALSE, fixed.

13 years agostylecontext: add a raised style class
Cosimo Cecchi [Tue, 8 Mar 2011 22:08:03 +0000 (17:08 -0500)]
stylecontext: add a raised style class

This is useful for raised elements, for example raised buttons on
primary toolbars.

https://bugzilla.gnome.org/show_bug.cgi?id=644248

13 years agoRemove support for GDK_NATIVE_WINDOWS
Alexander Larsson [Mon, 7 Mar 2011 15:01:46 +0000 (16:01 +0100)]
Remove support for GDK_NATIVE_WINDOWS

GDK_NATIVE_WINDOWS was a way to keep some old apps running that did weird
things in gtk2. We should not have to carry this forwards in gtk 3.x.

We do however keep a g_warning() call reminding people of this fact to
ease debugging when they try to port their applications.

https://bugzilla.gnome.org/show_bug.cgi?id=644119

13 years agogtksettings: plug a leak
Carlos Garnacho [Mon, 7 Mar 2011 22:05:03 +0000 (23:05 +0100)]
gtksettings: plug a leak

Make the GValue take the created GString.

13 years agocssprovider: Reset the scope before skipping to the next rule after an error happens.
Carlos Garnacho [Mon, 7 Mar 2011 22:00:16 +0000 (23:00 +0100)]
cssprovider: Reset the scope before skipping to the next rule after an error happens.

This is so the selector token gets parsed with the correct GScanner scope/config
on all situations.

13 years agocssprovider: plug two more leak fixes
Carlos Garnacho [Mon, 7 Mar 2011 21:59:45 +0000 (22:59 +0100)]
cssprovider: plug two more leak fixes

13 years agocssprovider: return immediately if there's remaining data behind a parsed value
Carlos Garnacho [Mon, 7 Mar 2011 21:42:33 +0000 (22:42 +0100)]
cssprovider: return immediately if there's remaining data behind a parsed value

parse_value() could fill in a GError, but yet (incorrectly) return G_TOKEN_NONE,
having the GError leaked as a result.

https://bugzilla.gnome.org/show_bug.cgi?id=642604 , Reported by Felix Riemann.

13 years agoFinish making the GtkSettings object identical on all backends so that it doesn't...
John Ralls [Mon, 7 Mar 2011 20:28:20 +0000 (12:28 -0800)]
Finish making the GtkSettings object identical on all backends so that it doesn't assert on non-X11 backends

13 years agox11: Copy from actual window, not parent
Benjamin Otte [Mon, 7 Mar 2011 16:45:08 +0000 (17:45 +0100)]
x11: Copy from actual window, not parent

This code is a relic from GTK2 days and should not be necessary anymore,
as code now makes sure to only copy those parts of the window that are
not overlapped by parent windows.

By deleting it we fix potential issues with composited and translucent
windows copying the wrong data.

https://bugzilla.gnome.org/show_bug.cgi?id=643416

13 years agoBump version
Matthias Clasen [Mon, 7 Mar 2011 15:08:48 +0000 (10:08 -0500)]
Bump version

13 years agoUpdate NEWS
Matthias Clasen [Mon, 7 Mar 2011 13:24:27 +0000 (08:24 -0500)]
Update NEWS

13 years agotreeview: propagate the selected state from the row
Cosimo Cecchi [Mon, 7 Mar 2011 05:17:08 +0000 (00:17 -0500)]
treeview: propagate the selected state from the row

An expander in a row can't be selected on its own otherwise. This way
themes can override the expander colors when the row is selected.

https://bugzilla.gnome.org/show_bug.cgi?id=644089

13 years agoentry: Do not clobber state when rendering the default icons.
Carlos Garnacho [Mon, 7 Mar 2011 10:59:33 +0000 (11:59 +0100)]
entry: Do not clobber state when rendering the default icons.

Use lower level methods that just require changing the state in the
GtkStyleContext instead. Fixes a problem where the insensitive state
was being set to the entry becasuse a parent was insensitive, spotted
by Alex Larsson.

13 years agonotebook: flip the render order
Cosimo Cecchi [Sat, 5 Mar 2011 08:01:44 +0000 (03:01 -0500)]
notebook: flip the render order

The render order for tabs is now

- left to right until the active tab
- right to left until the active tab
- active tab

This allows themes that use non-straight lines for the tab curvature to
draw them not worrying about flipping one side after the active tab.

13 years agonotebook: fixup the last commit
Cosimo Cecchi [Mon, 7 Mar 2011 03:04:13 +0000 (22:04 -0500)]
notebook: fixup the last commit

Let the tab overlap eat the padding, otherwise having tab-overlap >
tab-curvature without cutting the label is impossible.
This way we give the label widget all the allocation possible before
giving up and cutting it if the values for tab-overlap are too extreme
for the given padding.

13 years agonotebook: make sure the label widget allocation not exceeds the overlap
Cosimo Cecchi [Sat, 5 Mar 2011 08:05:12 +0000 (03:05 -0500)]
notebook: make sure the label widget allocation not exceeds the overlap

13 years agonotebook: make sure the allocation keeps track of the tab curvature
Cosimo Cecchi [Sat, 5 Mar 2011 08:05:54 +0000 (03:05 -0500)]
notebook: make sure the allocation keeps track of the tab curvature

This was a genuine typo, because "padding" was calculated but unused in
this branch of the code.

13 years agoUpdated Esperanto translation
Kristjan SCHMIDT [Sun, 6 Mar 2011 17:27:14 +0000 (18:27 +0100)]
Updated Esperanto translation

13 years agoUpdated Portuguese translation
Duarte Loreto [Sun, 6 Mar 2011 15:20:37 +0000 (15:20 +0000)]
Updated Portuguese translation

13 years agoUpdated Spanish translation
Jorge González [Sun, 6 Mar 2011 14:21:41 +0000 (15:21 +0100)]
Updated Spanish translation

13 years agoxi2: Ignore button release events of button 4-7
Carlos Garcia Campos [Sat, 5 Mar 2011 11:13:19 +0000 (12:13 +0100)]
xi2: Ignore button release events of button 4-7

We treat button presses of button 4-7 as scroll wheel events, so release
events should be ignored.

https://bugzilla.gnome.org/show_bug.cgi?id=643925

13 years agoFix conversion from GdkColor to GdkRGBA
Ignacio Casal Quinteiro [Sun, 6 Mar 2011 10:16:28 +0000 (11:16 +0100)]
Fix conversion from GdkColor to GdkRGBA

13 years agoUse existing mime type for GTK_FILE stock icon
Matthias Clasen [Sun, 6 Mar 2011 05:31:20 +0000 (00:31 -0500)]
Use existing mime type for GTK_FILE stock icon

document-x-generic is not actually a mimetype. Using text-x-generic
instead lets us work out-of-the-box in a11y themes.

13 years agoOnly count visible children during size request
Matthias Clasen [Sun, 6 Mar 2011 04:25:35 +0000 (23:25 -0500)]
Only count visible children during size request

This was causing spacing between image and label to be requested
in buttons, even though the image is invisible.

13 years agotheme: Fix spinner for inactive state.
Benjamin Otte [Sat, 5 Mar 2011 14:01:19 +0000 (15:01 +0100)]
theme: Fix spinner for inactive state.

Previously, the inactive state caused the spinner's solid line to be
drawn to the right. Now it's drawn to the top instead, which makes the
inactive state look more natural.

13 years agotheme: Make spinner animation turn clockwise
Benjamin Otte [Sat, 5 Mar 2011 13:59:25 +0000 (14:59 +0100)]
theme: Make spinner animation turn clockwise

13 years agotheme: Make spinner steps count forward, not backward
Benjamin Otte [Sat, 5 Mar 2011 13:57:55 +0000 (14:57 +0100)]
theme: Make spinner steps count forward, not backward

This makes the spinner roate the right way again.

13 years agotheme: Make spinner drawing start with the solid line
Benjamin Otte [Sat, 5 Mar 2011 13:53:26 +0000 (14:53 +0100)]
theme: Make spinner drawing start with the solid line

Previously, for i == 0, we were drawing the translucent line first.
However, people reading the code would always assume the solid line is
indicated by the current step. So make the code do that.

Note that the spinner now rotates in the wrong direction. This will be
fixed in a future commit.

13 years agotheme: Compute scale values once
Benjamin Otte [Sat, 5 Mar 2011 13:49:19 +0000 (14:49 +0100)]
theme: Compute scale values once

Use a local variable to store the xscale and yscale, so that it's only
computed once.
Makes it easier to do the changes I'm about to do.

13 years agotheme: Do general cairo setup in one place
Benjamin Otte [Sat, 5 Mar 2011 13:47:55 +0000 (14:47 +0100)]
theme: Do general cairo setup in one place

Also remove wrong comment.

13 years agotheme: Remove unnecessary save/restore in spinner code
Benjamin Otte [Sat, 5 Mar 2011 13:46:36 +0000 (14:46 +0100)]
theme: Remove unnecessary save/restore in spinner code

13 years agotheme: Simplify spinner drawing code a bit.
Benjamin Otte [Sat, 5 Mar 2011 13:30:17 +0000 (14:30 +0100)]
theme: Simplify spinner drawing code a bit.

13 years agogtk_init*(): Document that argc/argv can be 0,NULL.
Murray Cumming [Sat, 5 Mar 2011 13:53:52 +0000 (14:53 +0100)]
gtk_init*(): Document that argc/argv can be 0,NULL.

* gtk/gtkmain.c: (gtk_init): Mention that argc and argv can be 0,
  using the new text from the g_applicatoin_run() documentation.
  There was already introspection annotation about that.
  (gtk_init_check, gtk_init_with_args): Use the same text here.
  See bug #643649 comment 2.

13 years agoAdded Latvian translation.
Rudolfs Mazurs [Sat, 5 Mar 2011 13:48:26 +0000 (15:48 +0200)]
Added Latvian translation.

13 years agoAdded Spanish translation
Daniel Mustieles [Sat, 5 Mar 2011 12:16:06 +0000 (13:16 +0100)]
Added Spanish translation

13 years agoUpdated Portuguese translation
Duarte Loreto [Fri, 4 Mar 2011 23:34:51 +0000 (23:34 +0000)]
Updated Portuguese translation

13 years agoUpdated Portuguese translation
Duarte Loreto [Fri, 4 Mar 2011 23:33:55 +0000 (23:33 +0000)]
Updated Portuguese translation

13 years agoUpdated Portuguese translation
Duarte Loreto [Fri, 4 Mar 2011 23:32:59 +0000 (23:32 +0000)]
Updated Portuguese translation

13 years agospinbutton: remove unused variable
Xan Lopez [Fri, 4 Mar 2011 20:43:08 +0000 (21:43 +0100)]
spinbutton: remove unused variable

13 years agoPrime NEWS for 3.0.2
Matthias Clasen [Fri, 4 Mar 2011 18:55:06 +0000 (13:55 -0500)]
Prime NEWS for 3.0.2

13 years agospinbutton: rework GtkSpinButton drawing handler
Cosimo Cecchi [Mon, 28 Feb 2011 20:44:53 +0000 (15:44 -0500)]
spinbutton: rework GtkSpinButton drawing handler

Make the arrows bigger, and position them according to the mockups.

13 years agoRevert "spinbutton: center-align by default"
Cosimo Cecchi [Fri, 4 Mar 2011 16:12:32 +0000 (11:12 -0500)]
Revert "spinbutton: center-align by default"

This reverts commit 64ef11985bb11de3e1369cfa3d187d274096cf6e.

13 years agoAdded UG translation
Abduxukur Abdurixit [Fri, 4 Mar 2011 15:36:06 +0000 (16:36 +0100)]
Added UG translation

13 years agoAdded UG translation
Abduxukur Abdurixit [Fri, 4 Mar 2011 15:29:30 +0000 (16:29 +0100)]
Added UG translation

13 years agoAdded UG translation
Abduxukur Abdurixit [Fri, 4 Mar 2011 15:27:22 +0000 (16:27 +0100)]
Added UG translation

13 years agoUploaded Ukranian
Daniel Korostil [Fri, 4 Mar 2011 08:58:08 +0000 (10:58 +0200)]
Uploaded Ukranian

13 years agoUpdated Polish translation
Piotr Drąg [Fri, 4 Mar 2011 08:54:03 +0000 (09:54 +0100)]
Updated Polish translation

13 years agoBump the Pango requirement to 1.24
Matthias Clasen [Fri, 4 Mar 2011 00:24:02 +0000 (19:24 -0500)]
Bump the Pango requirement to 1.24

This avoids warnings in some situations. 1.24 is 2 years old,
so should not cause any problems.
https://bugzilla.gnome.org/show_bug.cgi?id=643370

13 years agoGtkNotebooks: fix regressions from tab-pack removal
Matthias Clasen [Thu, 3 Mar 2011 23:26:09 +0000 (18:26 -0500)]
GtkNotebooks: fix regressions from tab-pack removal

13 years agoGtkNotebook: Don't let tab allocation go negative
Matthias Clasen [Thu, 3 Mar 2011 23:24:58 +0000 (18:24 -0500)]
GtkNotebook: Don't let tab allocation go negative

13 years agoCleanup unused variables
Cosimo Cecchi [Thu, 3 Mar 2011 22:47:50 +0000 (17:47 -0500)]
Cleanup unused variables

13 years agodocs: update to include the new style classes
Cosimo Cecchi [Thu, 3 Mar 2011 15:13:29 +0000 (10:13 -0500)]
docs: update to include the new style classes

13 years agostylecontext: add a primary toolbar style class
Cosimo Cecchi [Thu, 3 Mar 2011 15:13:10 +0000 (10:13 -0500)]
stylecontext: add a primary toolbar style class

13 years agoall: add a "separator" style class
Cosimo Cecchi [Thu, 3 Mar 2011 03:55:25 +0000 (22:55 -0500)]
all: add a "separator" style class

13 years agomenuitem: set the menubar style class for menu items on a bar
Cosimo Cecchi [Wed, 2 Mar 2011 23:41:24 +0000 (18:41 -0500)]
menuitem: set the menubar style class for menu items on a bar

So that themes can distinguish them from the other menu items.

13 years agowindow: don't confuse priv->type_hint with priv->gdk_type_hint
Cosimo Cecchi [Wed, 2 Mar 2011 21:06:09 +0000 (16:06 -0500)]
window: don't confuse priv->type_hint with priv->gdk_type_hint

This broke, among the others, menu rendering.

13 years agomenuitem: mark with the GTK_STYLE_CLASS_MENUITEM style class
Cosimo Cecchi [Wed, 2 Mar 2011 03:33:15 +0000 (22:33 -0500)]
menuitem: mark with the GTK_STYLE_CLASS_MENUITEM style class

13 years agoUse 16 as default size for checkbutton indicator
Cosimo Cecchi [Wed, 2 Mar 2011 03:31:13 +0000 (22:31 -0500)]
Use 16 as default size for checkbutton indicator

So that 16x16 icons can be used from themes such as Adwaita in their
natural size.

13 years agoswitch: remove slider padding from available allocation when drawing
Cosimo Cecchi [Tue, 1 Mar 2011 14:00:19 +0000 (09:00 -0500)]
switch: remove slider padding from available allocation when drawing

This also simplifies the code a bit, and makes previous workarounds
useless.

13 years agoswitch: fetch the padding values from the slider
Cosimo Cecchi [Tue, 1 Mar 2011 13:59:02 +0000 (08:59 -0500)]
switch: fetch the padding values from the slider

It's the slider which acts as a child inside the container through, so
we should fetch the slider padding when computing height/width.

13 years agoswitch: propagate the active state to the slider
Cosimo Cecchi [Mon, 28 Feb 2011 23:46:46 +0000 (18:46 -0500)]
switch: propagate the active state to the slider

13 years agocombobox: add an arrow-scaling property to GtkComboBox
Cosimo Cecchi [Mon, 28 Feb 2011 21:37:52 +0000 (16:37 -0500)]
combobox: add an arrow-scaling property to GtkComboBox

13 years agospinbutton: center-align by default
Cosimo Cecchi [Mon, 28 Feb 2011 20:45:51 +0000 (15:45 -0500)]
spinbutton: center-align by default

13 years agoarrow: x and y coordinates of the arrow rendering should be double
Cosimo Cecchi [Sun, 27 Feb 2011 00:29:44 +0000 (19:29 -0500)]
arrow: x and y coordinates of the arrow rendering should be double

To prevent off-by-one rounding errors when drawing them later.

13 years agorange: x and y coordinates of the arrow rendering should be double
Cosimo Cecchi [Sun, 27 Feb 2011 00:29:08 +0000 (19:29 -0500)]
range: x and y coordinates of the arrow rendering should be double

To prevent off-by-one rounding errors when drawing them later.

13 years agorange: don't set junction sides on scrollbar steppers
Cosimo Cecchi [Sun, 27 Feb 2011 00:28:33 +0000 (19:28 -0500)]
range: don't set junction sides on scrollbar steppers

13 years agoGtkScale: Add style classes for marks
Matthias Clasen [Thu, 3 Mar 2011 21:47:55 +0000 (16:47 -0500)]
GtkScale: Add style classes for marks

This enables better styling of the slider in the presence of
marks. Based on a patch by Bastien Nocera,
https://bugzilla.gnome.org/show_bug.cgi?id=643685

13 years agoGtkScale: Normalize mark positions
Matthias Clasen [Thu, 3 Mar 2011 21:30:21 +0000 (16:30 -0500)]
GtkScale: Normalize mark positions

This is in preparation for making scale marks styling better.
As a side-effect, TOP and LEFT position are treated the same now,
as well as BOTTOM and RIGHT.

Based on a patch by Bastien Nocera,
https://bugzilla.gnome.org/show_bug.cgi?id=643685

13 years ago[gi] don't skip gtk_menu_attach_to_widget, correctly annotate the async callback
John (J5) Palmieri [Thu, 3 Mar 2011 21:06:28 +0000 (16:06 -0500)]
[gi] don't skip gtk_menu_attach_to_widget, correctly annotate the async callback

13 years ago[gi] add missing annotations for signals that emit Gdk.Events
John (J5) Palmieri [Thu, 3 Mar 2011 21:05:28 +0000 (16:05 -0500)]
[gi] add missing annotations for signals that emit Gdk.Events

13 years agoAdd a test for interpretation of mark positions
Matthias Clasen [Thu, 3 Mar 2011 18:48:34 +0000 (13:48 -0500)]
Add a test for interpretation of mark positions

13 years agoDon't distribute gdkconfig.h
Matthias Clasen [Thu, 3 Mar 2011 03:17:02 +0000 (22:17 -0500)]
Don't distribute gdkconfig.h

https://bugzilla.gnome.org/show_bug.cgi?id=643440

13 years agoAdd GtkBuilder custom attributes for style classes
Matthias Clasen [Thu, 3 Mar 2011 01:07:36 +0000 (20:07 -0500)]
Add GtkBuilder custom attributes for style classes

GtkWidget now parses custom attributes like

  <style>
     <class name="dark-label"/>
     <class name="big-heading"/>
  </style>

to add style classes to widgets.

https://bugzilla.gnome.org/show_bug.cgi?id=643347

13 years agoUpdated Portuguese translation
Duarte Loreto [Wed, 2 Mar 2011 23:48:34 +0000 (23:48 +0000)]
Updated Portuguese translation

13 years agowindow: Add check for X11 window when calling x11 function
Benjamin Otte [Wed, 2 Mar 2011 16:29:21 +0000 (17:29 +0100)]
window: Add check for X11 window when calling x11 function

13 years agosettings: Always use the same default values
Benjamin Otte [Wed, 2 Mar 2011 16:18:49 +0000 (17:18 +0100)]
settings: Always use the same default values

Set the Mac key theme when creating a Quartz settings object instead of
having it the default when quartz is enabled.
This keeps compatibility with the GTK2 behavior that the Mac key theme
is not used for the X11 backend, which could now happen for a
multi-backend build.

13 years agosettings: Keep the GtkSettings object identical no matter the backend
Benjamin Otte [Wed, 2 Mar 2011 16:02:51 +0000 (17:02 +0100)]
settings: Keep the GtkSettings object identical no matter the backend

Previously we were enabling some settings properties only if the X11
backend was enabled. This worked fine with GTK2 where only one backend
was enabled at a time, but now when multiple backends can be enabled,
this does not make sense.

13 years agosettings: Avoid a segfault when loading keyfile
Benjamin Otte [Wed, 2 Mar 2011 15:21:44 +0000 (16:21 +0100)]
settings: Avoid a segfault when loading keyfile

When the keyfile is loaded, no screen exists yet, so we need to vheck
for that.

13 years agoAllow GtkAppChooser classes to be instantiated with content-type=NULL.
Tristan Van Berkom [Tue, 1 Mar 2011 10:41:15 +0000 (19:41 +0900)]
Allow GtkAppChooser classes to be instantiated with content-type=NULL.

Mostly for corner cases such as Glade that treats objects generically
and needs to have an object that doesnt crash out of g_object_new().

13 years agoupdate Russian translation
Leonid Kanter [Tue, 1 Mar 2011 14:20:21 +0000 (16:20 +0200)]
update Russian translation

13 years agoDon't use GDK_NONE when GDK_SELECTION_CLIPBOARD is meant
Matthias Clasen [Tue, 1 Mar 2011 00:24:29 +0000 (19:24 -0500)]
Don't use GDK_NONE when GDK_SELECTION_CLIPBOARD is meant

13 years agoRemoved incorrect description's sentence.
Patricia Santana Cruz [Mon, 28 Feb 2011 16:54:31 +0000 (17:54 +0100)]
Removed incorrect description's sentence.

Description was wrong when it explained: "The two areas are separated
by a GtkHSeparator.". These separators are not used any more.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=643496

13 years agogdk: add precondition checks to many GdkDisplay functions
Michael Natterer [Mon, 28 Feb 2011 16:34:52 +0000 (17:34 +0100)]
gdk: add precondition checks to many GdkDisplay functions

instead of crashing in GDK_DISPLAY_GET_CLASS(NULL)->foo(NULL);

13 years agoDon't assert on zero width/height in render functions
Matthias Clasen [Mon, 28 Feb 2011 12:52:59 +0000 (07:52 -0500)]
Don't assert on zero width/height in render functions

We can just silently ignore those rendering calls.
https://bugzilla.gnome.org/show_bug.cgi?id=643041

13 years agogdk: gdk_x11_device_xi2_get_state(): move stuff to local scope
Michael Natterer [Sun, 27 Feb 2011 14:06:41 +0000 (15:06 +0100)]
gdk: gdk_x11_device_xi2_get_state(): move stuff to local scope

13 years agoUploaded Ukranian
Daniel Korostil [Sun, 27 Feb 2011 10:20:37 +0000 (12:20 +0200)]
Uploaded Ukranian

13 years agoDrop _gdk_x11_set_input_focus_safe
Matthias Clasen [Sun, 27 Feb 2011 03:13:06 +0000 (22:13 -0500)]
Drop _gdk_x11_set_input_focus_safe

All that this function did was to asynchronously catch errors.
Since we have asynchronous error traps now, we can just use
XSetInputFocus().

13 years agoUploaded Ukranian
Daniel Korostil [Sat, 26 Feb 2011 20:01:32 +0000 (22:01 +0200)]
Uploaded Ukranian

13 years agoList all available icons in Makefile.am
Matthias Clasen [Sat, 26 Feb 2011 05:23:47 +0000 (00:23 -0500)]
List all available icons in Makefile.am

This mismatch was causing e.g. gtk-caps-lock-warning.png to show
up as missing image in entries.

13 years agoGtkSwitch: use regular activation
Matthias Clasen [Fri, 25 Feb 2011 20:47:03 +0000 (15:47 -0500)]
GtkSwitch: use regular activation

GtkSwitch was listening for Space/Enter keyreleases itself, instead
of providing an 'activate' action signal, like e.g GtkButton. As a
side-effect, this fixes
https://bugzilla.gnome.org/show_bug.cgi?id=643321

13 years agoGtkTrayIcon: chain up in dispose
Matthias Clasen [Fri, 25 Feb 2011 19:37:57 +0000 (14:37 -0500)]
GtkTrayIcon: chain up in dispose

Pointed out in bug 643131

13 years agoMore patchwork for icon view subclasses that access the icon view api prematurely.
Tristan Van Berkom [Fri, 25 Feb 2011 07:00:20 +0000 (16:00 +0900)]
More patchwork for icon view subclasses that access the icon view api prematurely.

In this patch we adress rows_reordered() and row_deleted() callbacks
(since some icon view subclasses manipulate the connected treemodel
from _init()).

13 years agoMore patchwork to ensure GtkIconView subclasses can use the icon view api
Tristan Van Berkom [Fri, 25 Feb 2011 05:53:12 +0000 (14:53 +0900)]
More patchwork to ensure GtkIconView subclasses can use the icon view api
from the subclass's instance structure initializer

13 years agoCopy event axes for double/triple click events in gdk_event_copy()
Carlos Garcia Campos [Thu, 24 Feb 2011 16:38:40 +0000 (17:38 +0100)]
Copy event axes for double/triple click events in gdk_event_copy()

Button event axes for double/triple click events are freed by
gdk_event_free(), so copy them to avoid an invalid free when releasing a
copied event.

13 years ago[GtkFileChooser] Fix initial selection when hidden files are shown
Federico Mena Quintero [Thu, 24 Feb 2011 20:53:09 +0000 (14:53 -0600)]
[GtkFileChooser] Fix initial selection when hidden files are shown

We were breaking prematurely out of the loop that goes through
each file that is pending selection; not only does that loop
disable filtering and show hidden files if needed, but it also
selects the files themselves.  So we need to walk the whole
list of files.

Signed-off-by: Federico Mena Quintero <federico@gnome.org>
https://bugzilla.gnome.org/show_bug.cgi?id=643170

13 years agoAdded documentation to "cell-area" construct-only properties.
Tristan Van Berkom [Thu, 24 Feb 2011 07:05:41 +0000 (16:05 +0900)]
Added documentation to "cell-area" construct-only properties.

Indicate what kind of area will be used by default if none is
provided by the user.

13 years agoAvoid spurious emissions of monitors-changed
Alexander Larsson [Thu, 24 Feb 2011 18:44:16 +0000 (19:44 +0100)]
Avoid spurious emissions of monitors-changed

The monitor change detection code in _gdk_x11_screen_size_changed() and
process_monitors_change() goes to some length to make sure its only emitted
when there is an actual change to the data visible via the GdkScreen monitors
api.

However, commit 662e69ad added some code that always emits "monitors-changed"
in _gdk_x11_screen_size_changed when we have randr13 and get a ConfigureNotify
on the root window (even though we may already have emitted it in the
RRScreenChangesNotify event!).

As far as I can tell this is due to a comment in the bug referenced by the
commit (https://bugzilla.gnome.org/show_bug.cgi?id=601712#c4) where it says:

  This version of the patch changes GdkDisplay to emit "monitors-changed" when
  the primary monitor changes (see the change in _gdk_x11_screen_size_changed).

And, if you remove this part of the change the signal is not emitted when just
the primary is changed. However, this is not really the right approach. We
should just also check for if the primary changes in process_monitors_change()
to avoid spurious signal emissions.

https://bugzilla.gnome.org/show_bug.cgi?id=643216

13 years agoUpdated Slovenian translation
Klemen Košir [Thu, 24 Feb 2011 16:23:23 +0000 (17:23 +0100)]
Updated Slovenian translation

13 years agoRemoved unused EXPOSE_EVENT enumeration from the GtkWidget signals.
Tristan Van Berkom [Wed, 23 Feb 2011 13:07:27 +0000 (22:07 +0900)]
Removed unused EXPOSE_EVENT enumeration from the GtkWidget signals.

13 years agoRemove mention of GtkWidget::expose-event from docs in gtkwidget.c
Tristan Van Berkom [Wed, 23 Feb 2011 12:58:48 +0000 (21:58 +0900)]
Remove mention of GtkWidget::expose-event from docs in gtkwidget.c