]> dgit.raspbian.org Git - gnome-shell.git/log
gnome-shell.git
9 years agooverview: Don't expand workspace thumbnails by default
Florian Müllner [Thu, 1 Jun 2017 13:38:41 +0000 (15:38 +0200)]
overview: Don't expand workspace thumbnails by default

We currently expand the workspace switcher when workspaces are being
used, that is when there are any windows on a non-active workspace.
While this helps with the switcher's discoverability, it does eat into
the space available for window previews. By now the component should
be well established, so we can afford opting for space efficiency and
only expand the switcher while the user actually interacts with it.

https://bugzilla.gnome.org/show_bug.cgi?id=783953

9 years agoworkspaceThumbnails: Reduce maximum thumbnail size
Florian Müllner [Thu, 1 Jun 2017 13:49:46 +0000 (15:49 +0200)]
workspaceThumbnails: Reduce maximum thumbnail size

The overview's window picker is primarily about windows, and as the
previews that represent them are more effective the bigger they are,
it makes sense to scale down competing elements; start by reducing
the size of workspace thumbnails on the right ...

https://bugzilla.gnome.org/show_bug.cgi?id=783953

9 years agogdmUtils: Define constants with 'var'
Florian Müllner [Thu, 10 Aug 2017 16:35:07 +0000 (18:35 +0200)]
gdmUtils: Define constants with 'var'

Most of those are accessed outside the module itself, but commit
033277b6 missed them ...

https://bugzilla.gnome.org/show_bug.cgi?id=785556

9 years agoworkspaceSwitcherPopover: Remove pointless timeout
Florian Müllner [Fri, 21 Jul 2017 23:08:44 +0000 (01:08 +0200)]
workspaceSwitcherPopover: Remove pointless timeout

Ever since commit a4e53953, the popup has been hidden initially, so
there's little point in setting up a timeout handler to hide it ...

9 years agoUpdate Hungarian translation
Balázs Meskó [Thu, 10 Aug 2017 08:14:44 +0000 (08:14 +0000)]
Update Hungarian translation

9 years agoUpdate Swedish translation
Anders Jonsson [Wed, 9 Aug 2017 20:20:00 +0000 (20:20 +0000)]
Update Swedish translation

9 years agoUpdate Indonesian translation
Andika Triwidada [Wed, 9 Aug 2017 05:08:32 +0000 (05:08 +0000)]
Update Indonesian translation

9 years agoUpdate Friulian translation
Fabio Tomat [Tue, 8 Aug 2017 19:01:26 +0000 (19:01 +0000)]
Update Friulian translation

9 years agoUpdated Slovenian translation
Matej Urbančič [Mon, 7 Aug 2017 08:07:11 +0000 (10:07 +0200)]
Updated Slovenian translation

9 years agoUpdate Galician translation
Fran Dieguez [Mon, 7 Aug 2017 08:05:10 +0000 (08:05 +0000)]
Update Galician translation

9 years agoUpdated Slovenian translation
Matej Urbančič [Mon, 7 Aug 2017 06:49:17 +0000 (08:49 +0200)]
Updated Slovenian translation

9 years agoUpdate Catalan translation
Jordi Mas [Sat, 5 Aug 2017 08:13:49 +0000 (10:13 +0200)]
Update Catalan translation

9 years agoUpdated Spanish translation
Daniel Mustieles [Sat, 5 Aug 2017 10:01:20 +0000 (12:01 +0200)]
Updated Spanish translation

9 years agoUpdate Slovak translation
Dušan Kazik [Fri, 4 Aug 2017 20:22:41 +0000 (20:22 +0000)]
Update Slovak translation

9 years agopadOsd: Add one to the >1 rings
Carlos Garnacho [Thu, 3 Aug 2017 19:59:26 +0000 (21:59 +0200)]
padOsd: Add one to the >1 rings

Ring leader/label SVG elements for the second ring are named *Ring2*.
We need to add one here for elements >1, this is consistent with strips.

https://bugzilla.gnome.org/show_bug.cgi?id=785358

9 years agoUpdated Czech translation
Marek Cernocky [Thu, 3 Aug 2017 09:23:40 +0000 (11:23 +0200)]
Updated Czech translation

9 years agoUpdate Croatian translation
gogo [Wed, 2 Aug 2017 19:51:59 +0000 (19:51 +0000)]
Update Croatian translation

9 years agoUse HTTPS URLs for the submodules
Mathieu Bridon [Wed, 2 Aug 2017 14:25:44 +0000 (16:25 +0200)]
Use HTTPS URLs for the submodules

The git protocol is usually blocked in conference/hotel networks, and it
is vulnerable to MITM attacks.

9 years agoUpdate POTFILES.in
Piotr Drąg [Wed, 2 Aug 2017 13:17:35 +0000 (15:17 +0200)]
Update POTFILES.in

9 years agoUpdated Czech translation
Marek Cernocky [Wed, 2 Aug 2017 11:52:27 +0000 (13:52 +0200)]
Updated Czech translation

9 years agoui: Add InhibitShortcutsDialog
Florian Müllner [Fri, 14 Jul 2017 01:15:17 +0000 (03:15 +0200)]
ui: Add InhibitShortcutsDialog

https://bugzilla.gnome.org/show_bug.cgi?id=783342

9 years agobuild: Bump mutter API version
Florian Müllner [Tue, 1 Aug 2017 15:40:52 +0000 (16:40 +0100)]
build: Bump mutter API version

Mutter bumped its API version to make all the new goodness
parallel-installable with current stable ones, adjust to
that.

9 years agoUpdate Friulian translation
Fabio Tomat [Tue, 1 Aug 2017 11:11:06 +0000 (11:11 +0000)]
Update Friulian translation

9 years agobackground: use of CHANGES_DONE_HINT events to avoid reloads
Cosimo Cecchi [Mon, 31 Jul 2017 15:17:35 +0000 (16:17 +0100)]
background: use of CHANGES_DONE_HINT events to avoid reloads

Instead of potentially loading a background mid-way when it changes, and
loading it again for every file monitor event, leverage
CHANGES_DONE_HINT events, which allow us to ignore CREATED and CHANGED
signals from the file monitor.

https://bugzilla.gnome.org/show_bug.cgi?id=747794

9 years agowindowManager: use Meta.MonitorManager.get_display_configuration_timeout()
Cosimo Cecchi [Sat, 18 Jun 2016 00:14:02 +0000 (17:14 -0700)]
windowManager: use Meta.MonitorManager.get_display_configuration_timeout()

Instead of hardcoding a 20 seconds limit.

https://bugzilla.gnome.org/show_bug.cgi?id=767805

9 years agoUpdated Slovenian translation
Matej Urbančič [Fri, 28 Jul 2017 11:04:35 +0000 (13:04 +0200)]
Updated Slovenian translation

9 years agoUpdate Kazakh translation
Baurzhan Muftakhidinov [Thu, 27 Jul 2017 12:10:41 +0000 (12:10 +0000)]
Update Kazakh translation

9 years agoUpdated Spanish translation
Daniel Mustieles [Tue, 25 Jul 2017 18:15:15 +0000 (20:15 +0200)]
Updated Spanish translation

9 years agoUpdate Croatian translation
gogo [Tue, 25 Jul 2017 18:07:47 +0000 (18:07 +0000)]
Update Croatian translation

9 years agoUpdate Friulian translation
Fabio Tomat [Tue, 25 Jul 2017 13:43:59 +0000 (13:43 +0000)]
Update Friulian translation

9 years agoUpdated Lithuanian translation
Aurimas Černius [Mon, 24 Jul 2017 20:46:14 +0000 (23:46 +0300)]
Updated Lithuanian translation

9 years agoUpdate Nepali translation
Pawan Chitrakar [Mon, 24 Jul 2017 11:30:13 +0000 (11:30 +0000)]
Update Nepali translation

9 years agokeyboard: Only rebuild keyboard actor on keyboard type changes
Carlos Garnacho [Sun, 23 Jul 2017 11:51:11 +0000 (13:51 +0200)]
keyboard: Only rebuild keyboard actor on keyboard type changes

About every other situation can do with synchronizing keyboard visibility,
and keyboard layout changes are already handled internally in the Keyboard
object.

A downside of this approach is that once created, there will always be a
Keyboard instance and its full actor hierarchy. Seems reasonable to do that
since we can't tell it won't ever be needed.

https://bugzilla.gnome.org/show_bug.cgi?id=785309

9 years agokeyboard: Remove dead code
Carlos Garnacho [Sun, 23 Jul 2017 11:40:31 +0000 (13:40 +0200)]
keyboard: Remove dead code

There's no Caribou daemon proxy anymore, no need to shut it down either.

https://bugzilla.gnome.org/show_bug.cgi?id=785309

9 years agoUpdate Friulian translation
Fabio Tomat [Sat, 22 Jul 2017 21:01:29 +0000 (21:01 +0000)]
Update Friulian translation

9 years agoworkspace: Switch back to using the :has-pointer property
Florian Müllner [Sat, 22 Jul 2017 09:47:20 +0000 (11:47 +0200)]
workspace: Switch back to using the :has-pointer property

In case where a method- and property name overlap, using the method
is less unambiguous than I thought - mozjs52-based gjs will only see
the method, while mozjs38-based gjs will only see the property. We
are in luck though, and the real property name contains dashes that
allow us to refer to the property in a way that works for all gjs
versions.

https://bugzilla.gnome.org/show_bug.cgi?id=785090

9 years agoRevert "networkAgent: Adjust to introspection change"
Emmanuele Bassi [Fri, 21 Jul 2017 14:34:50 +0000 (15:34 +0100)]
Revert "networkAgent: Adjust to introspection change"

This reverts commit 4181035981720ea0be863846f2524b767a46d22c.

The commit that caused this change was reverted in GLib.

9 years agonetworkAgent: Adjust to introspection change
Florian Müllner [Thu, 20 Jul 2017 15:14:31 +0000 (17:14 +0200)]
networkAgent: Adjust to introspection change

GLib commit fd329f4853f1 added annotations to GKeyFile methods that
change the API on the JS level, so adjust to that.

9 years agobuild: Fix portal installation with meson
Florian Müllner [Wed, 19 Jul 2017 23:43:38 +0000 (01:43 +0200)]
build: Fix portal installation with meson

Gah, this fell through the cracks ...

9 years agobuild: Update gvc subproject
Florian Müllner [Wed, 19 Jul 2017 23:28:35 +0000 (01:28 +0200)]
build: Update gvc subproject

9 years agoBump version to 3.25.4
Florian Müllner [Wed, 19 Jul 2017 22:20:23 +0000 (00:20 +0200)]
Bump version to 3.25.4

Update NEWS.

9 years agobuild: Support the meson build system
Florian Müllner [Thu, 25 May 2017 00:16:27 +0000 (02:16 +0200)]
build: Support the meson build system

Meson is on track to replace autotools as the build system of choice,
so support it in addition to autotools. If all goes well, we'll
eventually be able to drop the latter ...

https://bugzilla.gnome.org/show_bug.cgi?id=783229

9 years agodocs: Use entity for version
Florian Müllner [Sun, 28 May 2017 19:46:02 +0000 (21:46 +0200)]
docs: Use entity for version

Meson's gtk-doc support currently requires the main SGML/XML file
to be located in the srcdir, so instead of substituting the module
version in that file directly, use a small auxiliary file that's
included via a custom entity.

https://bugzilla.gnome.org/show_bug.cgi?id=783229

9 years agobuild: Move gvc submodule to /subprojects/gvc
Florian Müllner [Thu, 25 May 2017 16:28:03 +0000 (18:28 +0200)]
build: Move gvc submodule to /subprojects/gvc

When using the meson build system, gvc will be built as a subproject.
As meson restricts submodules to a subprojects/ directory in the
source tree root, move the submodule there.

https://bugzilla.gnome.org/show_bug.cgi?id=783229

9 years agoworkspace: Use has_pointer() method instead of property
Florian Müllner [Wed, 19 Jul 2017 00:43:25 +0000 (02:43 +0200)]
workspace: Use has_pointer() method instead of property

ClutterActor has both a has_pointer() method and a :has-pointer
property (that we represent as 'has_pointer'). So far gjs was
able to deal with the name overlap, but now trying to use the
property will instead test for the availability of the method.
Just avoid the conflict by switching to the method, which is
unambiguous.

https://bugzilla.gnome.org/show_bug.cgi?id=785090

9 years agoiconGrid: Cancel animations on unmap
Florian Müllner [Mon, 17 Jul 2017 19:58:47 +0000 (21:58 +0200)]
iconGrid: Cancel animations on unmap

We now cancel animations on override, however we also want to cancel
animations altogether on unmap (that is, when hiding the overview)
to avoid icons swarming into the void.

https://bugzilla.gnome.org/show_bug.cgi?id=736148

9 years agoiconGrid: Remove clones on new animation
Carlos Soriano [Fri, 5 Sep 2014 17:33:37 +0000 (19:33 +0200)]
iconGrid: Remove clones on new animation

Until now we were waiting the animation to complete to allow the user to
make a new animation. This could bring some problems and annoy nervous
users.
Instead of that, destroy clones on new animations triggers and
create a new animation with the new direction.

https://bugzilla.gnome.org/show_bug.cgi?id=736148

9 years agoappDisplay: Use ::paint handler to set up spring animation
Florian Müllner [Fri, 2 Jun 2017 02:22:36 +0000 (04:22 +0200)]
appDisplay: Use ::paint handler to set up spring animation

The animation needs the icons' final positions, so we currently defer
it to a ::notify::allocation handler; however as starting the animation
during an allocation cycle would trigger a Clutter warning, it is
further deferred to a MetaLater. While this usually works, it is possible
that the allocation is already valid when we connect the signal, in which
case the animation is triggered at a later unexpected time. Switch to
a more robust ::paint handler instead, which also allows us to get rid
of the double-delay.

https://bugzilla.gnome.org/show_bug.cgi?id=736148

9 years agoenvironment: Remove stray semicolon
Florian Müllner [Mon, 17 Jul 2017 10:20:59 +0000 (12:20 +0200)]
environment: Remove stray semicolon

9 years agoDefine externally accessible contants with 'var' instead of 'const'
Florian Müllner [Tue, 18 Jul 2017 17:47:27 +0000 (19:47 +0200)]
Define externally accessible contants with 'var' instead of 'const'

Just as we did with classes, define other constants that are (or
may be) used from other modules with 'var' to cut down on warnings.

https://bugzilla.gnome.org/show_bug.cgi?id=785084

9 years agoDefine classes with 'var' instead of 'const'
Florian Müllner [Tue, 18 Jul 2017 17:41:25 +0000 (19:41 +0200)]
Define classes with 'var' instead of 'const'

Any symbols (including class properties) that should be visible
outside the module it's defined in need to be defined as global.
For now gjs still allows the access for 'const', but get rid of
the warnings spill now by changing it.

https://bugzilla.gnome.org/show_bug.cgi?id=785084

9 years agomain: Define global properties with 'var' instead of 'let'
Florian Müllner [Tue, 18 Jul 2017 17:48:08 +0000 (19:48 +0200)]
main: Define global properties with 'var' instead of 'let'

Symbols that are defined with 'let' are no longer visible outside
the module that defines them. To unbreak the code base, define all
non-private properties as global.

https://bugzilla.gnome.org/show_bug.cgi?id=785084

9 years agoappDisplay: Define constant before using it
Florian Müllner [Tue, 18 Jul 2017 17:23:12 +0000 (19:23 +0200)]
appDisplay: Define constant before using it

This is now a fatal error with current gjs, however I wonder: Did this
ever work?

https://bugzilla.gnome.org/show_bug.cgi?id=785084

9 years agomessageTray: Use spread syntax to get map keys
Florian Müllner [Tue, 18 Jul 2017 17:20:42 +0000 (19:20 +0200)]
messageTray: Use spread syntax to get map keys

It wasn't supported when the code was originally written, now the
fill-in code isn't accepted anymore ...

https://bugzilla.gnome.org/show_bug.cgi?id=785084

9 years agoFix crash when using the magnifier
Joaquim Rocha [Mon, 17 Jul 2017 10:55:20 +0000 (12:55 +0200)]
Fix crash when using the magnifier

The first time that the session is started, it can happen that the
AT SPI hasn't been correctly initialized, and this results in a crash
when attempting to register the caret or focus listeners.

In order to avoid this, these changes check the result of initializing
the AT SPI, to allow further attempts when it has failed.

https://bugzilla.gnome.org/show_bug.cgi?id=785047

9 years agopadOsd: Disallow help/switch monitor actions on rings/strips
Carlos Garnacho [Wed, 12 Jul 2017 16:37:18 +0000 (18:37 +0200)]
padOsd: Disallow help/switch monitor actions on rings/strips

It does not make sense to map those actions to non-buttons. Set the
actions insensitive in the combobox to disallow this from the UI.

https://bugzilla.gnome.org/show_bug.cgi?id=782033

9 years agopadOsd: Implement edition of actions for rings/strips
Carlos Garnacho [Wed, 5 Jul 2017 22:05:22 +0000 (00:05 +0200)]
padOsd: Implement edition of actions for rings/strips

Customization of keycombo actions for strips/rings was lost in the
porting to new incarnation of Wacom support.

The UI here is slightly different, instead of requiring the user to
rotate/swipe in each direction to map each keycombo, the UI will
navigate the user through edition of both options, first one, then
the other.

https://bugzilla.gnome.org/show_bug.cgi?id=782033

9 years agoRemove excess parameter to g_app_info_launch()
Florian Müllner [Sat, 15 Jul 2017 01:42:06 +0000 (03:42 +0200)]
Remove excess parameter to g_app_info_launch()

Commit 39a840e2c3d added an additional parameter to shell_app_launch().
When adjusting callers, the parameter was also added accidentally to
calls of the confusingly similar g_app_info_launch() ...
Remove those to fix some warnings.

https://bugzilla.gnome.org/show_bug.cgi?id=781471

9 years agoextensionSystem: Handle all initExtension() extension errors
Florian Müllner [Wed, 17 May 2017 22:03:28 +0000 (00:03 +0200)]
extensionSystem: Handle all initExtension() extension errors

The method currently catches errors that occur when calling the
extension's init() method, but throws itself an error if the
expected extension.js file is missing. The former is pointless
if we expect all callers to handle errors themselves anyway, and
we should avoid the latter if we don't - opt for the second option
and handle a missing extension.js file gracefully.

https://bugzilla.gnome.org/show_bug.cgi?id=781728

9 years agoextensionSystem: Catch import errors
Florian Müllner [Wed, 17 May 2017 21:57:18 +0000 (23:57 +0200)]
extensionSystem: Catch import errors

While we catch errors that occur when calling init(), enable() or
disable(), the import itself can throw an exception, for instance
if the extension imports an unavailable typelib or tries to draw
in a conflicting library.

https://bugzilla.gnome.org/show_bug.cgi?id=781728

9 years agoextensions: Consistently handle createExtensionObject() errors
Florian Müllner [Wed, 17 May 2017 21:51:49 +0000 (23:51 +0200)]
extensions: Consistently handle createExtensionObject() errors

The method may throw an error, for example when metadata.json is
missing or cannot be parsed, however we are currently not always
handling it.

https://bugzilla.gnome.org/show_bug.cgi?id=781728

9 years agoRevert "extensionSystem: handle reloading broken extensions"
Florian Müllner [Wed, 17 May 2017 21:27:37 +0000 (23:27 +0200)]
Revert "extensionSystem: handle reloading broken extensions"

Both reloadExtensions() and enableExtensions() are already expected
to catch extension errors. If they don't, this is the bug that
should be fixed instead of catching unhandled exceptions in the
caller.

This reverts commit ff425d1db7082e2755d2a405af53861552acf2a1.

https://bugzilla.gnome.org/show_bug.cgi?id=781728

9 years agoUpdate Friulian translation
Fabio Tomat [Mon, 17 Jul 2017 09:50:25 +0000 (09:50 +0000)]
Update Friulian translation

9 years agoUpdate Indonesian translation
Kukuh Syafaat [Mon, 17 Jul 2017 05:23:48 +0000 (05:23 +0000)]
Update Indonesian translation

9 years agopolkitAgent: Fix ReferenceError exception
Carlos Garnacho [Sun, 16 Jul 2017 19:24:30 +0000 (21:24 +0200)]
polkitAgent: Fix ReferenceError exception

Introduced in commit 593b4318a. The "message" variable was renamed
everywhere, except where it's defined.

9 years agoui: Implement "window doesn't respond" dialog on gnome-shell
Carlos Garnacho [Thu, 19 Jan 2017 18:52:18 +0000 (19:52 +0100)]
ui: Implement "window doesn't respond" dialog on gnome-shell

This does allow us to use ClutterActors instead of lousy legacy tools
meant for shell scripting.

https://bugzilla.gnome.org/show_bug.cgi?id=762083

9 years agoui: Use MessageDialogContent where appropriate
Florian Müllner [Sat, 15 Jul 2017 04:03:55 +0000 (06:03 +0200)]
ui: Use MessageDialogContent where appropriate

The gros of our ModalDialogs follow a UI pattern that matches the
newly added widget, so port them over to cut down on duplication.

https://bugzilla.gnome.org/show_bug.cgi?id=784985

9 years agoextensionDownloader: Don't reuse .prompt-* style classes
Florian Müllner [Sun, 16 Jul 2017 12:57:59 +0000 (14:57 +0200)]
extensionDownloader: Don't reuse .prompt-* style classes

Those will go away when we port authentication prompts to the new
MessageDialogContent widget, so pick the style classes from there
and adjust individual properties with more specific rules to re-
produce the existing style.

https://bugzilla.gnome.org/show_bug.cgi?id=784985

9 years agodialog: Add MessageDialogContent
Florian Müllner [Sat, 15 Jul 2017 03:12:57 +0000 (05:12 +0200)]
dialog: Add MessageDialogContent

A lot of our modal dialogs share a similar structure:

 [Icon]   Some title
          Maybe a subtitle

          And sometimes even a body for stuff like
          longer descriptions.

A dedicated widget with a common style will allow us to significantly
reduce duplication of both code and CSS.

https://bugzilla.gnome.org/show_bug.cgi?id=784985

9 years agoshellMountOperation: Share a style class
Florian Müllner [Sat, 15 Jul 2017 06:53:34 +0000 (08:53 +0200)]
shellMountOperation: Share a style class

Two of the dialogs use virtually identical styles, so it makes sense
to reuse the actual style classes.

https://bugzilla.gnome.org/show_bug.cgi?id=784985

9 years agoui: Refactor modal dialog into separate Dialog class
Carlos Garnacho [Thu, 19 Jan 2017 18:49:52 +0000 (19:49 +0100)]
ui: Refactor modal dialog into separate Dialog class

This is the basic dialog actor implementation, which will allow us to
use the same implementation on the session-global modal dialogs. The
ModalDialog class now uses it underneath, and so do all users of it.

https://bugzilla.gnome.org/show_bug.cgi?id=762083

9 years agolayout: Allow adding already parented actors for chrome tracking
Carlos Garnacho [Fri, 14 Jul 2017 20:23:53 +0000 (22:23 +0200)]
layout: Allow adding already parented actors for chrome tracking

let trackChrome accept actors that are not children of chrome actors.
this will be useful for the MetaCloseDialog in gnome-shell, which
is already included in the MetaWindowGroup, but needs to be tracked
as chrome for the dialog to receive pointer events on X11.

https://bugzilla.gnome.org/show_bug.cgi?id=762083

9 years agoslider: Cache slider width for dragging handlers
Carlos Garnacho [Fri, 14 Jul 2017 23:24:40 +0000 (01:24 +0200)]
slider: Cache slider width for dragging handlers

The actor allocation will be invalid on CLUTTER_TOUCH_BEGIN, because
it comes together with a CLUTTER_ENTER event that will recalculate
styles, and queue a relayout in result.

The net result is that on CLUTTER_TOUCH_BEGIN, the relayout has been
already queued, so the slider width comes up as 0, and the value ends
up as 1. Later touch events already happen on a validated actor, so
it is corrected. Still, not fun when modifying the volume slider on a
touchscreen.

9 years agoshell-util: Fix a typo
Alessandro Bono [Fri, 14 Jul 2017 21:38:52 +0000 (23:38 +0200)]
shell-util: Fix a typo

9 years agognome-shell: Make search result description semitransparent
Rares Visalom [Thu, 13 Jul 2017 19:02:20 +0000 (22:02 +0300)]
gnome-shell: Make search result description semitransparent

There is a clear difference between the file name and its associated
description which the mockups clearly highlight. Consequently,
make the description label semitransparent, in order to emphasize
the file name, not the description itself.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agosearch: Disable multiline descriptions
Rares Visalom [Sun, 9 Jul 2017 16:25:49 +0000 (19:25 +0300)]
search: Disable multiline descriptions

Having descriptions with multiple lines will clutter
the view and make it more confusing for the user. Apart
from that, it also makes the search result a lot bigger,
potentially losing general vertical alignment.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agosearch: Highlight search terms in description
Rares Visalom [Tue, 27 Jun 2017 22:01:13 +0000 (01:01 +0300)]
search: Highlight search terms in description

The classes extending the SearchResult can now connect
to the 'terms-changed' signal emitted by the SearchResult
class. This signal enables each object to update its
internal description in order to apply the bold style
onto strings that match the search terms.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agosearch: Propagate searchResults down the hierarchy
Rares Visalom [Tue, 27 Jun 2017 21:58:07 +0000 (00:58 +0300)]
search: Propagate searchResults down the hierarchy

In order to prepare for applying the bold style to the part
of the description that matches the search terms, we need
to listen for the signal that announces the fact that the
search terms have changed. Given the fact that the
SearchResults class is aware of the changes regarding the
search terms, the classes that extent SearchResult need
to have a reference to it in order to listen for the
to-be-implemented 'terms-changed' signal.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agoutil: Wrap g_regex_escape_string()
Florian Müllner [Fri, 23 Jun 2017 18:12:38 +0000 (20:12 +0200)]
util: Wrap g_regex_escape_string()

9 years agosearch: Increase maximum number of visible list search results
Rares Visalom [Fri, 23 Jun 2017 14:24:34 +0000 (17:24 +0300)]
search: Increase maximum number of visible list search results

Since results are now much more concise and take up less
space, we can use the 'saved' space to provide the user
with more search results for each provider.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agognome-shell: Use standard font size for the search-result-title class
Rares Visalom [Thu, 13 Jul 2017 19:01:09 +0000 (22:01 +0300)]
gnome-shell: Use standard font size for the search-result-title class

The current size of the text is way bigger than what
we want now. The default front size is what we are
looking after, therefore no explicit value is needed
to specify the font size.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agoseparator: Remove HorizontalSeparator
Rares Visalom [Thu, 13 Jul 2017 18:37:04 +0000 (21:37 +0300)]
separator: Remove HorizontalSeparator

It's now unused.

9 years agosearch: Replace the gradient separator with a simple line
Rares Visalom [Thu, 13 Jul 2017 18:34:47 +0000 (21:34 +0300)]
search: Replace the gradient separator with a simple line

The current separator uses a gradient effect as a separator
between search results. As the mockups suggest, the gradient
separator is no longer needed, in favor of a more simple one,
which is a thin semitransparent line.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agosearch: Tweak the sizes of search results icons
Rares Visalom [Tue, 20 Jun 2017 19:21:20 +0000 (22:21 +0300)]
search: Tweak the sizes of search results icons

Just a small tweak of icon sizes, in order to make
them look like in the mockup.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agosearch: Refactor providerIcon into providerInfo
Rares Visalom [Wed, 12 Jul 2017 23:51:59 +0000 (02:51 +0300)]
search: Refactor providerIcon into providerInfo

In order to match the current mockups, the providerIcon
class needed to include both the name of the provider
and the label that informs the user about how many more
search results are available for that specific provider.
The latter replaces the plus sign icon that has been
used so far.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agosearch: Make the results take more horizontal space
Rares Visalom [Tue, 20 Jun 2017 19:13:48 +0000 (22:13 +0300)]
search: Make the results take more horizontal space

In order to make gnome-shell search functionality fit on
smaller screens, like those of devices, search results
need to take advantage of more horizontal space so that
any extra space can be used efficiently.

In order to do so, change the layout of the ListSearchResult
class from a vertical one, to a horizontal one and also
decrease the padding of the list-search-result-content css
class.

https://bugzilla.gnome.org/show_bug.cgi?id=749957

9 years agopadOsd: strengthen ring/strip label creation
Carlos Garnacho [Wed, 12 Jul 2017 20:38:31 +0000 (22:38 +0200)]
padOsd: strengthen ring/strip label creation

If the padOsd is given a nonexistent ring/strip, things would fail
badly later when trying to paint a 0x0 StLabel. Just avoid creating
more ring/strip labels than those known by libwacom.

This is unlikely to happen, but seems better to protect against it.

https://bugzilla.gnome.org/show_bug.cgi?id=782033

9 years agopadOsd: Refactor function setting label after edition into separate function
Carlos Garnacho [Wed, 5 Jul 2017 22:03:12 +0000 (00:03 +0200)]
padOsd: Refactor function setting label after edition into separate function

https://bugzilla.gnome.org/show_bug.cgi?id=782033

9 years agoui: Reset workspace thumbnail porthole on monitor changes
Shih-Yuan Lee (FourDollars) [Fri, 9 Jun 2017 10:26:11 +0000 (18:26 +0800)]
ui: Reset workspace thumbnail porthole on monitor changes

The porthole will not be destroyed when the scale factor changed.
That makes workspace thumbnail porthole still wrong size in the first
seeing after the scale factor changed.

https://bugzilla.gnome.org/show_bug.cgi?id=765011

9 years agoshell-util: ensure we've a valid n_captures when capturing images
Marco Trevisan (Treviño) [Thu, 8 Jun 2017 12:16:04 +0000 (20:16 +0800)]
shell-util: ensure we've a valid n_captures when capturing images

https://bugzilla.gnome.org/show_bug.cgi?id=765011

9 years agoshell-util: Fix capturing scaled views
Jonas Ådahl [Wed, 7 Jun 2017 06:53:25 +0000 (14:53 +0800)]
shell-util: Fix capturing scaled views

libmutter-clutter fixed stage capturing when views are scaled, and for
that to work properly when compositing the screenshot, we need to deal
with the cairo device scale here too.

https://bugzilla.gnome.org/show_bug.cgi?id=765011

9 years agoshell-global: Don't set the clutter font DPI
Jonas Ådahl [Thu, 11 May 2017 04:22:27 +0000 (12:22 +0800)]
shell-global: Don't set the clutter font DPI

The clutter font DPI is managed by libmutter.

https://bugzilla.gnome.org/show_bug.cgi?id=765011

9 years agoshell-global: Don't override the X11 window scale
Jonas Ådahl [Thu, 11 May 2017 04:21:16 +0000 (12:21 +0800)]
shell-global: Don't override the X11 window scale

It's already done by libmutter.

https://bugzilla.gnome.org/show_bug.cgi?id=765011

9 years agost-icon: Remove unused fields
Jonas Ådahl [Thu, 4 May 2017 07:45:41 +0000 (15:45 +0800)]
st-icon: Remove unused fields

They were set but then never accessed.

https://bugzilla.gnome.org/show_bug.cgi?id=765011

9 years agoshellDBus: Avoid access to undefined properties
Florian Müllner [Sat, 17 Jun 2017 16:31:34 +0000 (18:31 +0200)]
shellDBus: Avoid access to undefined properties

Don't rely on an implicit undefined value of unset parameters to
shut up some warnings.

https://bugzilla.gnome.org/show_bug.cgi?id=781471

9 years agodateMenu: Set _delegate on non-menu-item menu child
Florian Müllner [Sat, 17 Jun 2017 16:14:20 +0000 (18:14 +0200)]
dateMenu: Set _delegate on non-menu-item menu child

PopupMenu needs special-treatment of certain types of menu items,
which it determines via children's _delegate property. However as
the calendar drop-down is very unmenu-ish, we use regular actors
rather than PopupMenuItems and the missing _delegate property
triggers a warning. Just add it as the bare minimum to make
PopupMenu happy.

https://bugzilla.gnome.org/show_bug.cgi?id=781471

9 years agonotificationDaemon: Avoid access to undefined property
Florian Müllner [Tue, 13 Jun 2017 04:49:07 +0000 (06:49 +0200)]
notificationDaemon: Avoid access to undefined property

Avoid a warning when assigning from an undefined property by making
sure to pass a proper null value instead.

https://bugzilla.gnome.org/show_bug.cgi?id=781471

9 years agocalendar: Fix access to undefined property
Florian Müllner [Tue, 13 Jun 2017 04:41:08 +0000 (06:41 +0200)]
calendar: Fix access to undefined property

Don't call a parent method before chaining-up _init() to
avoid a warning about access to an undefined property.

https://bugzilla.gnome.org/show_bug.cgi?id=781471

9 years agowindowManager: Fix a warning
Florian Müllner [Tue, 13 Jun 2017 03:51:31 +0000 (05:51 +0200)]
windowManager: Fix a warning

The destroy effect currently relies on a monkey-patched property
added from the map effect. However on X11 it is possible that we
did no map animation for a window that is destroyed when the shell
was restarted or had taken over from another WM. Just use the real
MetaWindow property to avoid a warning in that case.

https://bugzilla.gnome.org/show_bug.cgi?id=781471

9 years agolayout: Skip strut computation in the no-monitor case
Florian Müllner [Thu, 13 Jul 2017 13:53:04 +0000 (15:53 +0200)]
layout: Skip strut computation in the no-monitor case

It's possible for updateRegions() to be called before monitors have
been properly initialized. Instead of throwing an error in that case,
just skip the strut computation (that doesn't make sense anyway without
a monitor).

https://bugzilla.gnome.org/show_bug.cgi?id=781471