Alexander Larsson [Thu, 9 Feb 2012 22:16:03 +0000 (23:16 +0100)]
Remove all default css and make Raleigh a pure fallback theme
It was problematic to maintain Raleigh going forward, as any
changes in it affected all themes. Also, its more robust if
each theme is a full standalone css rather than relying on
an inherited css base.
So, this changes Raleigh to a standalone theme that we can tweak
without accidentally breaking other themes, and makes the
default theme empty. In fact, we don't even add the default
provider anymore as its always empty.
Piotr Drąg [Thu, 9 Feb 2012 20:32:56 +0000 (21:32 +0100)]
Updated POTFILES.skip
Alexander Larsson [Thu, 9 Feb 2012 20:09:44 +0000 (21:09 +0100)]
Fix transparency handling with non-double-buffered drawing
Sometimes we need to read back the window content into our double
buffer due to rendering a window with alpha when there is
no implicit paint or it has been flushed due to non-db drawing
before.
However, in this case we can't use gdk_cairo_set_source_window as
it might trigger an implicit paint flush as we detect what we
think is a direct non-double buffered window draw operation, which
will flush the implicit paint operation that we're just setting up.
To fix this we use the raw gdk_window_ref_impl_surface operation
to get the source surface.
Alexander Larsson [Thu, 9 Feb 2012 20:05:35 +0000 (21:05 +0100)]
Fix non-double-buffered drawing
There was a sign issue in a coordinate transform that made us
flush the wrong region when flushing an implicit paint.
The non-double buffered drawing would then be drawn over the
right area, but then at the end of the implicit paint this
would be overdrawn with the area we didn't properly remove
from the implicit paint.
Also, the translation from window coords to impl window
coords is now done before removing any active double
buffered paints, as these are also in impl window coords.
Kalev Lember [Thu, 9 Feb 2012 13:12:22 +0000 (15:12 +0200)]
win32: Don't crash when installed in a top-level directory
Avoid NULL-pointer dereference when package installation directory
doesn't contain any slashes.
Reported by Paweł Forysiuk.
Alexander Larsson [Thu, 9 Feb 2012 15:38:54 +0000 (16:38 +0100)]
Make the default background for GdkWindows transparent
With the changes in default CSS to make the default background transparent
we ran into issues where intermediate GdkWindow (for instance the
view_window in GtkViewport) where we didn't set an explicit background
(because before they were always covered). So instead of showing throught
the transparent windows were showing the default backgroind of the intermediate
window (i.e. black).
With this change we also needed to fix GtkViewport, as it was previously
relying on the bin and view windows to cover widget->window so that the
border was not visible if shadow_type was NONE.
Alexander Larsson [Thu, 9 Feb 2012 15:36:46 +0000 (16:36 +0100)]
viewport: Set frame styleclass when getting border size during size allocation
Without this you don't get the right border/padding for widget->window
and the border shows up under the scrolled contents.
Matthias Clasen [Thu, 9 Feb 2012 14:10:50 +0000 (09:10 -0500)]
Set BUILT_SOURCES
Matthias Clasen [Thu, 9 Feb 2012 14:06:18 +0000 (09:06 -0500)]
widget-factory: add the logo
Matthias Clasen [Thu, 9 Feb 2012 13:45:40 +0000 (08:45 -0500)]
Move widget-factory to demos/
Lets install this, for the benefit of artists everywhere.
Matthias Clasen [Thu, 9 Feb 2012 13:05:47 +0000 (08:05 -0500)]
gtk-demo: Use preprocessing for resources
Strip blanks from ui files, and convert pngs into pixdata.
This is useful and serves as a test of these preprocessing
options at the same time.
Matthias Clasen [Thu, 9 Feb 2012 12:59:57 +0000 (07:59 -0500)]
gtk-demo: Use gdk-pixbuf resource api
This is nicer than manual stream manipulation.
Matthias Clasen [Thu, 9 Feb 2012 12:57:11 +0000 (07:57 -0500)]
widget-factory: add an about dialog
This is a first step towards making widget-factory an installed
demo.
Baurzhan Muftakhidinov [Thu, 9 Feb 2012 09:09:30 +0000 (15:09 +0600)]
[l10n] Updated Kazakh translation
Matthias Clasen [Wed, 8 Feb 2012 12:26:25 +0000 (07:26 -0500)]
Avoid a C99ism
Move variable declarations to the beginning of the block.
https://bugzilla.gnome.org/show_bug.cgi?id=669511
Diego Escalante Urrelo [Wed, 8 Feb 2012 08:32:07 +0000 (03:32 -0500)]
gtkbuilder-menus: translation-domain can be NULL
GtkBuilder defaults its translation-domain to NULL, which is a perfectly
fine value for d_gettext() and friends.
https://bugzilla.gnome.org/show_bug.cgi?id=669638
Diego Escalante Urrelo [Wed, 8 Feb 2012 08:27:39 +0000 (03:27 -0500)]
gtkactiongroup: clarify set_translation_domain docs
NULL is an actual useful value for
gtk_action_group_set_translation_domain, but this information is buried
in g_dgettext() documentation.
Be extra redundant, to make it obvious that NULL is ok.
https://bugzilla.gnome.org/show_bug.cgi?id=669636
Mattias Põldaru [Wed, 8 Feb 2012 08:59:07 +0000 (10:59 +0200)]
[l10n] Updated Estonian translation
Matthias Clasen [Mon, 6 Feb 2012 22:32:46 +0000 (17:32 -0500)]
Bump version
Matthias Clasen [Mon, 6 Feb 2012 21:38:47 +0000 (16:38 -0500)]
3.3.14
Matthias Clasen [Mon, 6 Feb 2012 21:11:27 +0000 (16:11 -0500)]
Adapt expected output
A recent string change in GtkAboutDialog needs to be reflected
here.
Benjamin Otte [Mon, 6 Feb 2012 20:49:17 +0000 (21:49 +0100)]
tests: Remove untrue assertion
Even if something is not at the top, things might still not be scrolled.
For example in the case where things are exactly at the center.
Benjamin Otte [Mon, 6 Feb 2012 20:24:59 +0000 (21:24 +0100)]
tests: Don't test something we can't test
When using GtkTreeStore, don't compute row positions assuming
everything's a list store, which is how get_row_start_for_index() works.
So instead, do a laxer check and don't insist on proper alignment.
Benjamin Otte [Mon, 6 Feb 2012 20:21:09 +0000 (21:21 +0100)]
tests: Remove unused argument from treeview-scrolling test
Matthias Clasen [Mon, 6 Feb 2012 20:50:40 +0000 (15:50 -0500)]
Work around a drawing problem with small notebooks
Matthias Clasen [Mon, 6 Feb 2012 20:43:47 +0000 (15:43 -0500)]
More resetting
Matthias Clasen [Mon, 6 Feb 2012 20:09:24 +0000 (15:09 -0500)]
Use larger screen for tests
We used to run Xvfb with 800x600, which turns out to be too small
for some of our reftests. Bump it to 1024x768.
Matthias Clasen [Mon, 6 Feb 2012 19:20:20 +0000 (14:20 -0500)]
Fix declarations-valid-15 css parser test
Benjamin Otte [Mon, 6 Feb 2012 19:42:04 +0000 (20:42 +0100)]
reftests: Fix background-origin reftest for all themes
We want to @import reset-to-defaults.css here.
Benjamin Otte [Mon, 6 Feb 2012 18:50:11 +0000 (19:50 +0100)]
tests: Fix height checks in treeview scrolling test
If things get centered in the scrolling area depens on the height of the
row, too.
Benjamin Otte [Mon, 6 Feb 2012 18:27:45 +0000 (19:27 +0100)]
tests: Fix assertion emission
We want to assert() when doing the actual check, not do some
g_assert (testing_failed());
Matthias Clasen [Mon, 6 Feb 2012 19:06:20 +0000 (14:06 -0500)]
Fix declarations-valid-14 css parser test
Matthias Clasen [Mon, 6 Feb 2012 03:36:40 +0000 (22:36 -0500)]
Updates
Alexander Larsson [Mon, 6 Feb 2012 14:06:29 +0000 (15:06 +0100)]
Fix leak when parsing image urls
Swecha Localization Team [Mon, 6 Feb 2012 12:00:09 +0000 (17:30 +0530)]
Updated Telugu Translations
Piotr Drąg [Mon, 6 Feb 2012 01:38:25 +0000 (02:38 +0100)]
Add bloatpad to POTFILES.skip
It doesn't break the build yet, but it will once intltool is fixed to accept translatable='yes'
Matthias Clasen [Mon, 6 Feb 2012 00:12:41 +0000 (19:12 -0500)]
Mark strings as translable in bloatpad
We don't have any intention of actually translating bloatpad,
but this ends up as example in the docs.
Matthias Clasen [Sun, 5 Feb 2012 22:55:51 +0000 (17:55 -0500)]
Bump requirements of glib and gdk-pixbuf
Fran Diéguez [Sun, 5 Feb 2012 22:45:17 +0000 (23:45 +0100)]
Updated Galician translations
Kjartan Maraas [Sat, 4 Feb 2012 11:31:58 +0000 (12:31 +0100)]
Updated Norwegian bokmål translation
Daniel Mustieles [Fri, 3 Feb 2012 17:06:49 +0000 (18:06 +0100)]
Updated Spanish translation
Ignacio Casal Quinteiro [Fri, 3 Feb 2012 16:13:38 +0000 (17:13 +0100)]
aboutdialog: add a \n in the copyright text
Alexander Larsson [Fri, 3 Feb 2012 15:37:54 +0000 (16:37 +0100)]
Fix crash with theme resource file
The freeing the GResource in gtk_css_provider_reset caused problems
because it was called from gtk_css_provider_load_from_path() inside
gtk_css_provider_get_named(). We fix this by delaying the
setting of priv->resource until we successfully loaded the file.
Alexander Larsson [Fri, 3 Feb 2012 15:27:32 +0000 (16:27 +0100)]
Remove accidental leftover spew
Alexander Larsson [Fri, 3 Feb 2012 15:01:51 +0000 (16:01 +0100)]
css: Remove any loaded resource in gtk_css_provider_reset
Alexander Larsson [Fri, 3 Feb 2012 14:51:55 +0000 (15:51 +0100)]
Special case resource pixbuf files loading
This means we can share the pixbuf data when using GdkPixdata
images in the resource.
Alexander Larsson [Fri, 3 Feb 2012 14:50:49 +0000 (15:50 +0100)]
Support a gtk.gresource file in themes
This lets themes install a .gresource for efficient loading
of resources. This resource file will (if it exists) be automatically
loaded (mmaped) when the theme is used and unloaded when not used.
Piotr Drąg [Thu, 2 Feb 2012 16:25:40 +0000 (17:25 +0100)]
Updated POTFILES.skip
Daniel Mustieles [Thu, 2 Feb 2012 11:13:29 +0000 (12:13 +0100)]
Updated Spanish translation
Benjamin Otte [Thu, 2 Feb 2012 02:11:55 +0000 (03:11 +0100)]
reftests: Improve 'unresolvable' test
Include reset-to-defaults here, too
Benjamin Otte [Thu, 2 Feb 2012 01:58:59 +0000 (02:58 +0100)]
reftests: Fix css-match-exact
Ensure that background-colors are always set. And import
reset-to-defaults so that there's no background image or so.
Benjamin Otte [Thu, 2 Feb 2012 01:55:45 +0000 (02:55 +0100)]
reftests: Fix entry-progress-coloring test
Use reset-to-defaults here.
Benjamin Otte [Thu, 2 Feb 2012 01:55:05 +0000 (02:55 +0100)]
reftests: Add another background-size test
This one tests the keywords for the background-size property
Benjamin Otte [Tue, 24 Jan 2012 16:49:29 +0000 (17:49 +0100)]
css: Implement padding as numbers
Also remove the now unused border parsing function for shorthands.
Benjamin Otte [Tue, 17 Jan 2012 16:30:26 +0000 (17:30 +0100)]
reftests: Fix border-image-url reftest
"border-style: none;" sets border widths to 0. So no border was drawn.
Use "border-style: groove" instead. (If the border was drawn that way,
it'd look noticably different.)
Also include reset-to-defaults.css so we really get default CSS.
Benjamin Otte [Tue, 17 Jan 2012 16:12:32 +0000 (17:12 +0100)]
tests: Add 'px' to css tests
Benjamin Otte [Tue, 17 Jan 2012 16:09:30 +0000 (17:09 +0100)]
reftests: Add 'px' to all properties
Benjamin Otte [Tue, 17 Jan 2012 16:06:09 +0000 (17:06 +0100)]
reftests: Make test runner always use image surfaces
See comment in code for why.
Benjamin Otte [Tue, 17 Jan 2012 15:41:14 +0000 (16:41 +0100)]
css: Port margin properties to GtkCssNumber
Benjamin Otte [Tue, 17 Jan 2012 15:13:57 +0000 (16:13 +0100)]
reftests: Add reftest for background-size
Benjamin Otte [Wed, 18 Jan 2012 16:12:51 +0000 (17:12 +0100)]
reftests: Put border-image-repeat test back
This reverts commit
001d42f36955dd31f81e2e01a7b4dbb980a68b41.
Because now that we have background-size support, we can use it to make
stuff work!
Benjamin Otte [Tue, 17 Jan 2012 15:10:23 +0000 (16:10 +0100)]
reftests: set style properties to 0 in reset-to-defaults
GtkButton doesn't conform to CSS with these properties.
Bad GtkButton!
Benjamin Otte [Tue, 17 Jan 2012 15:09:35 +0000 (16:09 +0100)]
themingbackground: Implement background-size support
Benjamin Otte [Tue, 17 Jan 2012 14:15:34 +0000 (15:15 +0100)]
tests: Add parsing test for background-size
Benjamin Otte [Tue, 17 Jan 2012 14:11:48 +0000 (15:11 +0100)]
css: Implement parsing for 'background-size'
Benjamin Otte [Tue, 17 Jan 2012 11:14:50 +0000 (12:14 +0100)]
reftests: Add a bunch of linear gradient tests
Benjamin Otte [Tue, 17 Jan 2012 10:53:38 +0000 (11:53 +0100)]
tests: Add parsing tests for linear gradients
Benjamin Otte [Tue, 17 Jan 2012 10:44:07 +0000 (11:44 +0100)]
css: Implement repeating linear gradients
Benjamin Otte [Mon, 16 Jan 2012 17:28:58 +0000 (18:28 +0100)]
css: Implement CSS 3 linear-gradient() syntax
Benjamin Otte [Sun, 15 Jan 2012 18:47:38 +0000 (19:47 +0100)]
tests: Update CSS tests for borders
Borders now expect 'px'.
Benjamin Otte [Sun, 15 Jan 2012 16:51:58 +0000 (17:51 +0100)]
css: Implement support for angles
Benjamin Otte [Sun, 15 Jan 2012 05:26:08 +0000 (06:26 +0100)]
css: Convert border-radius to GtkCssNumber
We can do % now, wohoo!
Benjamin Otte [Sun, 15 Jan 2012 04:29:02 +0000 (05:29 +0100)]
themingengine: Add a private function to query lengths
Benjamin Otte [Sun, 15 Jan 2012 01:53:39 +0000 (02:53 +0100)]
css: Make border-width a length property
Benjamin Otte [Sun, 15 Jan 2012 01:30:26 +0000 (02:30 +0100)]
css: Add a check function for specified types
... and use it.
It seems kinda necessary for the refactoring I'm about to do...
Benjamin Otte [Sun, 15 Jan 2012 01:29:00 +0000 (02:29 +0100)]
css: Ensure proper type for parsing custom properties
Benjamin Otte [Fri, 13 Jan 2012 22:48:42 +0000 (23:48 +0100)]
css: Add GtkCssNumber
Benjamin Otte [Sat, 14 Jan 2012 16:20:15 +0000 (17:20 +0100)]
styleproperties: Add a warning
GtkStyleProperties stores computed values. Make sure it does.
Benjamin Otte [Sat, 14 Jan 2012 16:12:57 +0000 (17:12 +0100)]
styleproperty: Simplifiy registering function
... and rename it to conform to the actual type.
We don't need 2 registering functions anymore now that we can use the
specified type.
Benjamin Otte [Sat, 14 Jan 2012 16:04:00 +0000 (17:04 +0100)]
css: Allow bypassing the compute_value() stage
This is useful when overriding values.
Benjamin Otte [Sat, 14 Jan 2012 15:58:52 +0000 (16:58 +0100)]
csslookup: Query the bitmask
Checking if the value is NULL is the wrong thing to do - the bitmask is
usd to keep track of that.
The reason for that will become apparent in the next patch.
Benjamin Otte [Sat, 14 Jan 2012 04:10:46 +0000 (05:10 +0100)]
css: Remove two unused declarations
Benjamin Otte [Sat, 14 Jan 2012 04:02:25 +0000 (05:02 +0100)]
css: color_compute no longer needs to check type
The input is always a symbolic color.
Benjamin Otte [Sat, 14 Jan 2012 04:00:54 +0000 (05:00 +0100)]
css: Initialize value before calling compute function
Benjamin Otte [Sat, 14 Jan 2012 03:39:07 +0000 (04:39 +0100)]
css: Initialize values to their specified type
.. when parsing.
Benjamin Otte [Sat, 14 Jan 2012 03:18:29 +0000 (04:18 +0100)]
css: Create SymbolicColor fallback properties for colors
Benjamin Otte [Sat, 14 Jan 2012 03:05:19 +0000 (04:05 +0100)]
css: parse 'currentColor' for symbolic colors
Benjamin Otte [Sat, 14 Jan 2012 02:22:59 +0000 (03:22 +0100)]
styleproperty: Introduce "specified type" and "computed type"
Make the types explicit. This way, we can actually do useful stuff with
them (like sanity checks, d'oh).
Benjamin Otte [Sat, 14 Jan 2012 01:38:42 +0000 (02:38 +0100)]
css: Make the 'currentColor' keyword a symbolic color
This is the first step on a quest to ensure that there is only ever one
GType in use for GValues in every stage of the CSS resolving process.
Benjamin Otte [Wed, 1 Feb 2012 23:14:01 +0000 (00:14 +0100)]
raleigh: remove general 'padding' declaration
Fixes reftests that were broken after images and labels applied padding.
Things look much more cramped now. On the plus side, treeview headers
and comboboxes have better spacing. Menubars look like crap now though.
Needs someone with artistical talent to figure out.
Benjamin Otte [Wed, 1 Feb 2012 23:09:05 +0000 (00:09 +0100)]
raleigh: Remove unnecessary border-width declaration
Benjamin Otte [Wed, 1 Feb 2012 23:04:08 +0000 (00:04 +0100)]
raleigh: Don't set bg-color for all widgets
Just set oit for things that need it.
(FIXME: Buttons should probably not need it, but spinbuttons and
treeview headers don't have a proper background without it.)
Cosimo Cecchi [Wed, 1 Feb 2012 22:58:11 +0000 (17:58 -0500)]
reftests: set border-image: none in border-image-url's reference CSS
Fix border-image-url reftest
Cosimo Cecchi [Wed, 1 Feb 2012 22:57:45 +0000 (17:57 -0500)]
reftests: add a bunch of missing reset-to-defaults
Fix the css-match-* reftests.
Cosimo Cecchi [Wed, 1 Feb 2012 18:19:52 +0000 (13:19 -0500)]
app-chooser-dialog: cancel the PK operation when the dialog is dismissed
This doesn't really seem to have any effect; probably PK doesn't listen
for changes on the passed-in cancellable, but it's correct.
Related to: https://bugzilla.gnome.org/show_bug.cgi?id=649121
Cosimo Cecchi [Wed, 1 Feb 2012 18:16:06 +0000 (13:16 -0500)]
app-chooser-dialog: make sure to hold a ref when doing async ops
Hold a ref to the GtkDialog while doing async operations with the
GtkAppChooserOnline object.
This is needed, since somebody could call gtk_widget_destroy() on us
while an async operation is in progress. We don't want to be finalized
in that case, but mark the fact that we were dismissed and just return
from the callback in that case.
This avoids crashing if the dialog is destroyed in the middle of a PK
operation.
https://bugzilla.gnome.org/show_bug.cgi?id=649121
Benjamin Otte [Wed, 1 Feb 2012 14:44:27 +0000 (15:44 +0100)]
gtkmain: Invoke keysnoopers unconditionally
Now that ATK no longer uses a key snooper but is invoked directly,
checking in advance for existing snoopers is wrong and stops ATK from
working.
Also: code reduction without performance loss == good thing.
https://bugzilla.gnome.org/show_bug.cgi?id=669176
Cosimo Cecchi [Wed, 1 Feb 2012 01:18:09 +0000 (20:18 -0500)]
combobox: fix cell area allocation in menu mode for RTL
We were missing a padding to subtract from the child widget here.
Cosimo Cecchi [Wed, 1 Feb 2012 00:09:24 +0000 (19:09 -0500)]
combobox: fix dropdown menu alignment for RTL languages
We need to add in the left padding, instead of subtracting it in that
case.
Cosimo Cecchi [Wed, 1 Feb 2012 00:02:36 +0000 (19:02 -0500)]
combobox: use gtk_container_class_handle_border_width()
Avoids us fiddling with the border width manually in our size request
and allocation cycles.