Florian Müllner [Wed, 15 Feb 2012 16:59:50 +0000 (17:59 +0100)]
docDisplay: Remove "Recent Items" search provider
Although not all "Finding and reminding" applications are ready
yet, the integration with gnome-documents' search results overlaps
enough with the "Recent Items" provider to justify its removal.
Florian Müllner [Thu, 15 Dec 2011 13:46:12 +0000 (14:46 +0100)]
overview: Load RemoteSearchProviders
Allow applications to register search providers by dropping a keyfile
into a well-known directory. For now, initialize all found providers;
long term, we probably want to give users the ability to restrict the
set of active search providers.
Florian Müllner [Wed, 16 Nov 2011 23:11:05 +0000 (00:11 +0100)]
search: Add RemoteSearchProvider
Add an asynchronous search provider for results from a DBus service
implementing the org.gnome.Shell.SearchProvider interface; this
will allow applications to hook into the Shell's search without
implementing it in Shell itself or requiring an extension.
Florian Müllner [Mon, 6 Feb 2012 13:39:22 +0000 (14:39 +0100)]
search: Rename search_providers to open-search-providers
We will allow applications to hook into shell's search by registering
a service which implements a well-known DBus interface.
"search-providers" is a reasonable directory name for applications to
drop their registration files, but it conflicts with "search_providers"
used by open search providers - rename the latter to avoid confusion.
Florian Müllner [Tue, 21 Feb 2012 18:53:25 +0000 (19:53 +0100)]
search: Make asynchronous providers more explicit
Currently, asynchronous search providers are expected to call
startAsync() in getInitialResultSet()/getSubsearchResultSet(),
which will trigger async mode until the search is canceled or
updated. Switching between synchronous and asynchronous mode like
this makes asynchronous search an implementation detail, but being
transparent to the searchDisplay means that certain optimizations
don't work as expected. Namely, updating asynchronous search results
causes flickering, and the automatic selection never focuses
asynchronous results.
So change the API to require providers being either synchronous (with
the current getInitialResultSet()/getSubsearchResultSet() methods)
or asynchronous (with asynchronous variants), and handle asynchronous
providers explicitly in searchDisplay.
Florian Müllner [Fri, 17 Feb 2012 18:39:25 +0000 (19:39 +0100)]
searchDisplay: Split renderResults()
renderResults() updates the results set, determines the number of
results to display, retrieves the corresponding result metas and
adds a new results actor for each meta.
Splitting the function in those parts allows to move the retrieval
of the result metas into SearchResults, which is where we ensure
flicker-free rendering and control the selection - we want to keep
both features for asynchronous result metas which we are about to
introduce.
browser-plugin: Correct check for checking the hostname/protocol
While it's extremely unlikely that document.location would not be an
object in the browser setting, this check is incorrect and we could
possibly crash an NPAPI host if this is the case.
Florian Müllner [Mon, 20 Feb 2012 07:33:49 +0000 (08:33 +0100)]
app-system: Don't assume that gmenu_tree_load_sync() sets error
The function may return FALSE without setting the GError, so don't
assume it is set to prevent a crash in that case. While at it, free
the GError we were leaking before.
messageTray: Correctly show the icon on a notification stack
An actor is removed from its parent after it emits the destroy signal,
so we can't just check if the notification stack has more than one
notification -- we need to check if there's a notification there
that is not the current one.
This was causing spew in the form of:
"this.notificationStack.get_children()[0]._delegate.setIconVisible
is not a function"
Florian Müllner [Tue, 14 Feb 2012 15:30:40 +0000 (16:30 +0100)]
iconGrid: Split vertical/horizontal item size
IconGrid items used to be square, so a single size value made
perfect sense. However, as contact search uses rectangular items,
using a single size is wrong - the allocated height ends up twice
the size of the visible height, which is particularly visible if
another provider displays results below contact results.
Stefano Facchini [Wed, 15 Feb 2012 15:16:23 +0000 (16:16 +0100)]
main, lightbox: Fix lightbox for zoomed windows
The correct way to make an actor having the same size as another is
a ClutterBindConstraint. Connecting to 'allocation-changed' fails because
the allocation might not change even when 'width' and 'height' properties do.
This is the case of Main.uiGroup, used as parent container for zoomed
window clones.
In lightbox.js we bind also the position because in principle it could change,
even if currently only fullscreen lightboxes are used.
st-scroll-bar: Set the handle as a child of the bar, not the trough
The handle was a child of the trough, but it was allocated and painted
like it was a child of the bar. This will wreak havoc when we port over
to the new Clutter API, so let's just make it a child of the bar.
ShellNetworkAgent: don't access request fields if the operation is cancelled
When the operation is cancelled by NetworkManager, the request is
cancelled immediately. Later when gnome-keyring invokes the callback
notifying the error we must therefore not access its memory.
Previously the callback would mistakenly treat "cancelled" (which
indicates a programmatic cancel) as "denied" (which means the user
clicked "Cancel" on the keyring prompt)
VPN secrets are stored by the plugins, that provide separate
helpers for authentication. This commit adds the support for invoking
the binaries and pass them connection details.
For plugins that support it (as exposed by their keyfile), we invoke
them in "external-ui-mode" and expect a set of metadata about the
secrets which is used to build a shell styled dialog.
Substitutions generated by configure don't resolve prefixes, so
cannot be used for paths. Config already had localedir, and next
commit will need libexecdir and sysconfdir, so just bite the bullet
and move to sed.
ConsoleKit is being obsoleted by systemd. Accordingly port the CK logic
in the gnome-shell automount manager to systemd-logind APIs.
This makes use of systemd-logind's native C APIs which are much easier
to use than the D-Bus APIs in this case, and much faster too (since they
are synchronous and directly query the kernel for the information we
need). The dependency is compile time optional, and in order to be nice
to the Debian folks g-s compiled with this enabled fill automatically
fall back to CK support on systems lacking systemd.
ConsoleKit is obsoleted by systemd-logind. Accordingly, port
the current CK code to systemd. In order to be nice to
the Debian people fall back to CK if systemd is not found,
so that the code makes the best of whatever it runs on.
Gsm.GetRegistrationInfo and Cdma.GetServingSystem return a single argument
consisting of a tuple, not three separate arguments. This is
a regression from the GDBus port.
Stefano Facchini [Sat, 11 Feb 2012 10:14:43 +0000 (11:14 +0100)]
Return expected type from handleDragOver for Activities button actors
Currently they return 'undefined' instead of something meaningful,
e.g. DND.DragMotionResult.CONTINUE. This was unnoticed because none
of the ancestors of the Activities button actors do any drag handling.
The only visible issue are JS errors generated when dragging, for example,
a window thumbnail over the button, because the cursor cannot be set.
Stefano Facchini [Fri, 10 Feb 2012 19:18:35 +0000 (20:18 +0100)]
xdndHandler: prevent dummy actor from interfering with Hot Corner
Since the dummy actor occupies exactly the same area of the Hot Corner,
it can be erroneously picked during xdnd operations. Fix this by
hiding it from pick.
Owen W. Taylor [Tue, 7 Feb 2012 23:20:14 +0000 (18:20 -0500)]
Add -Wno-error=deprecated declarations
Even with --enable-compile-warnings=error, avoid erroring out on deprecations
for the moment, since we are hitting many Clutter deprecations and some are
hard to fix.
If both spacing and -shell-grid-item-size are 0, as they would be with nothing
setting them, we enter an infinite loop where we try to compute the layout.
Avoid the situation entirely by defaulting -shell-grid-item-size to a sane
value instead of 0.
extensionUtils: Create and allow access to a new "extension" object
The "extension" object is what I previously called the "helper" object.
It contains the extension importer object as well as the metadata object.
Things that were previously added on to the metadata (state, path, dir, etc.)
are now part of this new "extension" object.
With the new importer changes brought on by the extension prefs tool,
extensions are left without a way to import submodules at the global scope,
which would make them rely on techniques like:
var MySubModule;
function init(meta) {
MySubModule = meta.importer.mySubModule;
}
That is, there's now a lot more meaningless boilerplate that nobody wants
to write and nobody wants to reivew.
Let's solve this with a few clever hacks.
Allow extensions to get their current extension object with:
let extension = imports.misc.extensionUtils.getCurrentExtension();
As such, extensions can now get their own extension object before the
'init' method is called, so they can import submodules or do other things
at the module scope:
browser-plugin: Provide new APIs for launching extension preferences
Add two new APIs, "launchExtensionPrefs" to let SweetTooth let the user
launch the extension preferences tool directly from the browser. To allow
SweetTooth to check if an extension can be configured, add a new key to
the 'metadata', 'hasPrefs', which is returned by the GetExtensionInfo/
ListExtensions DBus methods.
projects / gnome-shell.git / log