AbsListView was relying on TOUCH_MODE_DOWN to signal that the item was
enabled, but onInterceptTouchEvent uses it for other purposes. Check
before dispatching the event.
Change-Id: I6babd3872cb51ebdba8e97a3b349e0cc08cd1279
We don't want to publish this, but for risk mitigation we are hiding it
rather than rewriting/expanding the FileBackupHelper to accomodate the
absolute-path use cases that the system uses internally.
Change-Id: I513c97ec54de8dd7d28b10868d447d94b82d4ec3
Addresses:
Issue #2550648: PackageManagerService setComponentEnabledSetting unconditionally
writes Settings xml
Issue #2549084: Make PackageManager.addPermission have async version
Also make the writing of settings when changing the preferred activities to use
the same async mechanism, and fiddle with thread priorities in the background
thread to go up to foreground priority when holding the lock to write settings
and a few other places. (At some point we should really clean this up to never
acquire the main lock while in the background.)
Change-Id: Ib2b7632543f6fb3f92a225518579f3b2d15e1413
This transaction can involve the transport having to query a remote backend
over the wire, so it can take a Long Time(tm). Make it main-thread-safe by
making it asynchronous, with the results passed as a callback to the invoker's
RestoreObserver. We also make the IRestoreObserver callback interface
properly oneway.
Bug #2550665
Bug #2549422
Change-Id: If18a233a0a3d54c7b55101715c9e6195b762c5a0
If the icon is precomposed or the icon IS null, set the touch icon url. This
ensures there will be a touch icon set if either precomposed or regular is
specified. Precomposed icons take precedence.
Bug: 2546984
Change-Id: Id9291006ecda8659b823ab9629eb98a6f8d45f5e
Bug: 2136464
When the bluez device is created, we get the onDeviceCreated signal.
We add it to our cache when that happens. We can have a device created
even when its not bonded - as a result of OPP. So use this cache to avoid
a DBUS call to Bluez.
Change-Id: I9465da7d72a12a6888128ff40ac1fe598cbae3c3
This change cleans up the JavaDoc in SearchableInfo,
removes trailing whitespace, fixes an if-statement
style violoation, and re-hides some things that were
exposed for Froyo that query deprecated or discouraged
attributes:
SearchableInfo.getSearchButtonText()
SearchableInfo.getLabelId()
SearchableInfo.getIconlId()
class SearchableInfo.ActionKeyInfo
SearchableInfo.findActionKey()
Bug http://b/issue?id=2553524
Change-Id: Iee5521c844f74137166730e523d29ba07ef1856b
Bug #2553187
Limit the max number of lines for each entry in the ResolveActivity to 2. Also
make sure to discard potential styling tags and newlines/space characters from
package name and activity labels.
Change-Id: Ibda3688267aa948b921164d3d3abc8c9236e61a2
I also realized the ROTATION_NN constants were misdocumented, they
actually align the opposite direction from the "orientation" angle
computed in the code (i.e. orientation=90 is left side up, ROTATION_90
is right side up). I just changed the comment here to keep the change
minimal; I'll update the code to be more consistent outside of this
release branch.
Bug: 2549875
Change-Id: I79c4286be582a21e730b25a8a24f2e1137c6e4cb
Use the content width to calculate the zoom overview
width. We used to use the minimum preferred width
to calculate the overview width as it is what we
use to define the viewport width. But some sites,
like cbsnews, engadget, the content width will be
slightly wider than the viewport width decided by
the min preferred width. The result is we can still
scroll a little bit in the overview mode. This issue
is magnified when we introduce the overscroll feature.
By using content width as zoom overview width, we
will have content fit in the overview mode. But
some sites, like
"http://www.sfjazz.org/concerts/2010/spring/artists/towner.php"
will be layout wider in the overview mode than it
is in the zoom in state. This is a tradeoff.
Fix http://b/issue?id=2551119
When receiving a message to delete the root layer, delay it until the next
new picture message, so that the layers only disappear when we start displaying
the new page, and not the moment the DOM on the old page is destroyed.
