This attribute was introduced in HC and is only partially supported. It should
actually also be used by the layout to be meaningful.
Bug 3223987
Change-Id: Idb47be319a236326f7624fd865466356c19bc342
InterfaceConfiguration changed to use InetAddress and stop with the string->int->string
conversions.
bug:2542681
Change-Id: I11c4954547333c43bb840fa0469ddde57b0d043b
Already exposed the new window flags and layout parameters,
just forgot to make this API visible at the same time.
Bug: 3049580
Change-Id: If8dc3568eb2806fa21881c31b9f879d6045ca890
Fixes a case where ViewRoot does not recompute its size properly.
This caused various offsets to get out of sync and the window would
draw improperly.
Change-Id: Id0abb37bbf27b60de2dca4077c21040ec84ef163
Bug 3186626
Restore some of the changes introduced in CL 78854.
Also added a few comments to the documentation.
Change-Id: I775be85c33ebe4a2384a40b21eec57f34fddcd33
protocol. The DHCP client and DHCP server classes are not included in this
commit: still working out some integration issues with ethernet & wifi services
Change-Id: I0c9cc48b7d960005b73ecb757c1fa66f0eb68471
This new class allows listeners to receive callbacks with elapsed time
that are sent on the same animation updates as other animators in the system.
It will allow simulations that go beyond the current animation system while
handling the actual animation timing system and ensuring that those
simulations sync up with other typical animations.
Change-Id: Iac91c39634218793f6598a7dec5ed71dc9630258
Also issue #3281400: Rotating a retained instance fragment leaks the fragment manager
And turn off fragment debug logging.
Change-Id: Ibdd7db82bb35618021bcba421ba92ced7cd691c2
DO NOT MERGE Fix bug 3124899 - timeout in CTS test
ExpandableListViewBasicTest
A certain sequence of events could cause AbsListView to enter an
infinite event loop. Only report reaching a scroll boundary to the
overscroll code path if the Scroller wants to move further, not if
we're done moving.
Change-Id: Ib61897f619e83b28f08b91e82f2b901cb1739e1a
SkBitmap::getPixels() can return NULL. The rest of the JNI Bitmap
code treats this NULL as if the SkBitmap has transparent black
pixels. Bitmap_writeToParcel now does the same.
Change-Id: I5e70b42b3d22a8aea898ce342e590000325bd0f9
If the dimensions of the original are sufficiently different from the
device's preferred dimensions, just don't restore the image. This
avoids bad letterboxing / clipping on e.g. phone <-> tablet data
migration.
The expansion/shrinkage ratios used here allow restores of saved
wallpaper images among HVGA devices, among WVGA variants, and
among tablets; but skip restoring wallpapers across those
categories (where severe clipping or letterboxing would occur).
Bug 3261863
Change-Id: I75e75d6401d18f1df10d75796ee04e21d2302cfa
There was a bug with LayoutTransitions where if the animations
were of different duration (such as in the current status bar clock),
it was possible to end up in a bad situation because the previous animation
might outlast the new animation, causing the opposite end result
from what was expected. The fix was to keep track of the current
add/remove animations and to cancel any running ones prior to starting
new ones.
Change-Id: I884ce33ce0671f6ba6153ee3951c8d14c5ed5714
SDP records are now registered with a dbus call so we don't have to wait
for initiating auto connections.
Also reduce time to connect other profiles case by 2 secs.
Change-Id: I8f0eab6a95d3bfaf11a8eb7495a024949639d7fc
