Set 'ppd_filename_set' and 'destination_set' always to FALSE
before requesting the destination and updated PPD file so we can
distinguish when we have all information needed for update
of the PpPPDOptionWidget.
https://bugzilla.redhat.com/show_bug.cgi?id=1211580
Two finger scroll toggle should be sensitive if at least one of the
devices support it. g-s-d is responsible to set edge-scroll for the
devices which don't support double touch.
https://bugzilla.gnome.org/show_bug.cgi?id=745601
We assume that the RR configuration is valid in various callbacks from
both our modal dialogs which doesn't hold if there's an hotplug while
a dialog is open (e.g. monitor is plugged/unplugged).
Closing the dialog in that case seems the right thing to do since it
would be operating on an outdated view of the world otherwise and
likely end up crashing.
https://bugzilla.gnome.org/show_bug.cgi?id=655041
GnomeRRMode is a boxed type which means that if stored as such in a
tree model, the model creates a copy to use internally. In addition,
it means that gtk_tree_model_get() will also get a copy which must be
freed by the caller which we were not doing.
In this case though, we don't need the copies since all the
GnomeRRModes that we use outlive the model so we can simplify things
by just storing a plain pointer in the model instead.
https://bugzilla.gnome.org/show_bug.cgi?id=655041
The section headings are useful to understand the content
of this panel. Sadly, orca does not currently read them
at all. We can fix this by setting up accessible relations
that help orca figure this out.
https://bugzilla.gnome.org/show_bug.cgi?id=746755
Tablets have not always an eraser (most of the generic tablets like Huion,
UC-Logic, etc... don't). We should not reject such tablets.
Commit 54849a9 (wacom: Only the stylus and eraser tools need to exist)
mentioned that we were not sure about eraser, and I think we should not
assume one either.
To do so, we simply ignore the eraser xinput node and rely on
libwacom to actually provide the eraser information.
If the stylus does not have the eraser tip, we may fall in the
LAYOUT_OTHER case. We have a picture of a generic Wacom pen with an
eraser, and the leaders linking the widget to the picture are scrambled.
To prevent that, gray out the eraser pressure slider so that we do not
break the layout.
https://bugzilla.gnome.org/show_bug.cgi?id=746117
Create the scrolled window with vertical scrolling turned off. Turn it
on when there are more than one items in the list.
Note that the immediate parent of the list is a GtkViewport, so we
need to explicitly look for the GtkScrolledWindow.
https://bugzilla.gnome.org/show_bug.cgi?id=745584
Every time we got the list of devices from the devices GtkListStore, we
were leaking the only element that wasn't a NetDevice, the proxy page.
Make sure to unref' it by hand if we're not going to add it to the list
(the elements of which will be unref'ed properly).
https://bugzilla.gnome.org/show_bug.cgi?id=746411
NetObjects and CcNetworkPanel aren't widgets, and hold references to
each other. Simplify things by having NetObject hold a pointer to
CcNetworkPanel (and the singletons NMRemoteSettings and NMClient)
instead, and clear their pointers when the original object is unref'ed.
https://bugzilla.gnome.org/show_bug.cgi?id=746411
When looking up a bar for a particular stream, special-case the input
and output bars/streams, as they will not be added to the "bars"
hashtable (which is, as per comment, only for applications and the sound
events bar).
https://bugzilla.gnome.org/show_bug.cgi?id=738869
This results in crashes when switching inputs or outputs, as other parts
of the code expect only application bars being added to that hashtable.
The bar was destroyed when switching outputs (as the stream is removed before
a new one being added) but with the bar pointing to freed memory.
This reverts commit 22fa7f1292.
https://bugzilla.gnome.org/show_bug.cgi?id=746227
This will fix the bar not showing up when we switch the default
orientation. When we changed the orientation, we went through
update_layout() which shows the frame at the end. _init() didn't do
that.
The airbrush tool has only one button, and the button mapping combo box
does not work.
We should not bail out if the widget has been removed, that means that
the stylus does not support the second button.
Check upfront the number of buttons to know if the top button parameter
is available.
https://bugzilla.gnome.org/show_bug.cgi?id=746263
We currently install two signal handlers for button-press-event and
selection-changed. As the former also calls the latter, the two interact
in non-obvious ways that can lead to two alerts being played at the same
time under certain circumstances.
Avoid this problem by setting activate-on-single-click on the tree view
and listening to the row-activated signal to drive previews and model
changes.
https://bugzilla.gnome.org/show_bug.cgi?id=671446
The language input chooser is being created and destroyed every time
it is opened and closed. This is noticeably slow and is hardly
necessary as the values it contains should not change.
To fix this, these changes reuse the same instance of the input chooser
(by hiding and showing it) instead of renewing it.
The input chooser is also now shown using gtk_dialog_run to avoid having
it destroyed from the default delete event.
These changes prevent warnings when unreferencing the default input
source, locale and back rows which are never instanced by some locales
(and thus null).
The language input chooser is being leaked. The reason for the leak
is a "circular reference": the input chooser keeps a GtkBuilder
instance that gets destroyed when the chooser is, itself, destroyed.
However, the input chooser is never destroyed because the GtkBuilder
instance keeps a reference to it.
These changes dismiss the GtkBuilder once it is no longer needed in the
input chooser which fixes the leak.
