If the panel was left open when the screen was turned off,
in some cases it might get stuck in an "open" state (the
panel's expanded height would be nonzero) although the
status bar window is in fact fully collapsed (due to
makeExpandedInvisible). The next time the user would go to
open the panel, things would be in an inconsistent state and
the panel wouldn't come down (on phones, the settings panel
would be attempted, but still nothing would happen).
This was easiest to reproduce on the keyguard (turn on
screen, pull down panel, turn off screen).
Bug: 7260868
Change-Id: Iec0000ba020e5a519eb5b4d42ac273b6689a18bd
Existing primary users were never being marked as complete,
causing things that relied on this (e.g. showing the quick settings panel)
to break.
Bug:7282088
Change-Id: I9c8622f3cd0fb99a44477946d3db22fa2cbbc6fc
The failures were caused by the implementation of doTextRunCursor
passing a too-small value for contextCount into the underlying
getTextRunAdvances call (it wasn't accounting for the start offset).
Thus, when getTextRunAdvances made a copy of the text for its cache key,
it was getting a partial copy.
This patch fixes the size so the cache key always has a full copy of the
text.
Change-Id: I57e3ac6de7aef0e1f1c7000dc3d653f9b0d623d2
Some security screens aren't currently calling userActivity(). As such,
they allow keyguard to timeout before the user has a chance to enter
the required information.
The fix uses a TextWatcher to look for changes in the input text
and call userActivity() appropriately.
bug 7291431
Change-Id: I6d7889cc01a4d6bdbefefc5af478e812c35b1a49
Reduce reliance on Service interface overrides, instead
steering clients to the DreamService-specific lifecycle
hooks:
onAttachedToWindow .. onDreamingStarted ..
onDreamingStopped .. onDetachedFromWindow
The old Dream.java is finally gone now too.
Bug: 7281802
Change-Id: Ib7802c3397fde60ad1132fa49831da182eef4d7a
Pro tem, we ignore wifi configuration data when restoring system settings.
This is not ideal, but it *does* mean we do not bounce wifi off and on
again during the extended restore process, which in turn means we don't
interfere with things like the Play Store's download of applications.
We do continue to back up wifi configuration, and will start using that
data again when the new implementation that restores AP configurations
without having to bounce wifi comes to pass.
Also, this CL fixes a longstanding bug in BackupDataInput.skipEntityData()
that was being reproduced reliably once settings restore was skipping
the wifi-related entities in the restore stream.
Bug 7249405
Change-Id: I61520a9a116b66ebdf95734d09d9afd46406df01
Create a BLUETOOTH_STACK permission to share between all users of
Bluetooth apk. Bluetooth Manager binds to apk of current user
bug 6925422
Change-Id: I2df5546e977c2d77bc66d97d12479d933826922a
This fixes an issue where the device would not lock immediately
when the user quickly pressed power off / power on even if
configured to do so. We were suppressing the screen off
and wake up broadcast in this case.
Now we make sure to always send at least one broadcast to
indicate the transition. We still collapse back-to-back
full cycle transitions though so as not to end up enqueuing
useless broadcasts.
Bug: 7061116
Change-Id: I7211c5fd963c271c2b0aceb4d2f746063c629079
Updated Roboto fonts fixing issue where Tcommaaccent and Tcedilla
characters were encoded improperly (their values were switched). This
patch addresses bug 7198855 and bug 7070912. Original was mistakenly
submitted to master but should be applied to jb-mr1-dev instead.
Also, this commit reverts the addition of RobotoCondensed-Light and its
italic, for which we don't have any build file or XML metadata support.
Change-Id: If1c592474f1f7199e98bb98edf4de6466b43fa51
Bug: 7280093
User serial number was not being bumped up from 0 on a freshly wiped device
due to the userlist not existing. This was resulting in the first 2 users
getting the same serial number, messing up C2DM message delivery. This issue
doesn't show up if you boot the device before adding a user, so it wasn't
discovered earlier.
Change-Id: I8a5d99b9ad2ccfb66a16cacac9548ba53f0be387
When we add a second type of tethering we unify the notifications
into a single generic type. This is done by canceling the first
and replacing it, but this flow was improperly canceling, so
you could end up with orphaned tethering icons.
bug:7283605
Change-Id: I1d136f51592b4326d48578cf67b69122e45d4984
Make sure we get the right initial status for wifi displays when a new
display route is first added to the MediaRouter.
Bug 7289746
Change-Id: Id41f704fc88021bd4cdf21e9f4b3c810775e1aea
The write-permission check must occur after any destination-table
rewriting, otherwise any application would be able to write to
any global setting, by supplying a fraudulent "system" namespace
in the uri, but with a key name that will be redirected to global.
Bug 7289965
Change-Id: I122098a64e40d14e00d3cb6608c50aeb74faf7ce
