The CTS tests weren't receiving the onTileAdded callbacks since the
shade was not in the listening state when tiles were added. This is
not possible for users (since tiles are added while the UI is
listening), but add a flushMessagesAndUnbind call to make sure the
tiles always receive onTileAdded.
Since now flushMessagesAndUnbind can be called before a real call
to setBindService(true), add a clause that will clear out its
immediate unbind state.
Change-Id: Idd13c432a4af5741813da9a9b1b292a47fd9443b
Fixes: 31147742
Before there was a jump-cut when a window that was occluding Keyguard
was going away, leading to an ugly flicker. To fix this, we do the
following.
- Always show windows with FLAG_SHOW_WHEN_LOCKED above lockscreen, even
if they don't "match" the currently occluding app (which is null in the
animation case)
- Move wallpaper behind last window that is not hidden by policy, so the
window doesn't get occluded by the wallpaper.
- Add a flag in the setOccluded call whether to animate or not. SystemUI
then plays a nice animation when it's set.
- Override the animation to always be the animation that happens when we
exit a window which is revealing the wallpaper behind, to make it
consistent with the home screen case.
Fixes: 30829255
Change-Id: Ib3fe20fc9003a0f9f291c974740f044ed8707e75
Add QS tiles to the backup list. Non-system tiles will get removed
since they won't be installed when restore happens.
Change-Id: Iccf6e773384c45bd4d1f10c21aa8af356b3920d2
Bug: 28782938
Fixes a bug that would re-open the remote input when
it was sent while HUN'd. Also fixes a bug that would
not properly clear a notification if swiped away and
at the same time HUN'd and having a reply active.
Change-Id: I3e8b5f81eaffd8e4b8602f3dbdd025229ba201fa
Fixes: 30454772
Fixes a bug that would re-open the remote input when
it was sent while HUN'd. Also fixes a bug that would
not properly clear a notification if swiped away and
at the same time HUN'd and having a reply active.
Change-Id: I3e8b5f81eaffd8e4b8602f3dbdd025229ba201fa
Fixes: 30454772
- To fix b/29320695, we moved the initial recents preloading to boot
completed. However, that event happens fairly late in the boot
sequence, and preloading can instead be done in response to user
unlocked.
Bug: 29879478
Change-Id: I881bd9c4c3b91d73df4947bfd2cc55eb6e615d15
We can't do it anymore with the slippery flag, so remove the entry
in the tuner.
We'll clean up the code in master.
Change-Id: Ib3dee0c947e7925b5847f2edc4801495c659dfa5
Fixes: 30891575
Also increases the pulse duration and switches timing to
elapsed realtime instead of wall time for robustness. Also
remove triggering for notification light for simplifying code.
Only noisy notifications are able to trigger the light anyway and
noisy notifications already trigger ambient on their own.
Change-Id: I4efbe530bb1a60cabaa46c2b04bcd5e12cdef680
Fixes: 31004875
Also increases the pulse duration and switches timing to
elapsed realtime instead of wall time for robustness. Also
remove triggering for notification light for simplifying code.
Only noisy notifications are able to trigger the light anyway and
noisy notifications already trigger ambient on their own.
Change-Id: I4efbe530bb1a60cabaa46c2b04bcd5e12cdef680
Fixes: 31004875
Apply a bunch of improvements to the acceleration classifier:
- When dragging from the edges, we get some delay without
movement between the DOWN and MOVE, which confounds the
classifier. Now discounts data segments where the delay
does not match the expected 16ms.
- The distance ratio did not compensate for differences
in sampling. If it does it's equivalent to the speed
ratio however. The distance ratio was removed and the
impact of the speed ratio score doubled.
- If we cannot calculate the ratio, no longer penalize
the traces for this.
Bug: 27405075
Change-Id: I067eb4d478593afbb20354e5c85a05353e2b4184
(cherry picked from commit 6b312a09fc)
Apply a bunch of improvements to the acceleration classifier:
- When dragging from the edges, we get some delay without
movement between the DOWN and MOVE, which confounds the
classifier. Now discounts data segments where the delay
does not match the expected 16ms.
- The distance ratio did not compensate for differences
in sampling. If it does it's equivalent to the speed
ratio however. The distance ratio was removed and the
impact of the speed ratio score doubled.
- If we cannot calculate the ratio, no longer penalize
the traces for this.
Bug: 27405075
Change-Id: I067eb4d478593afbb20354e5c85a05353e2b4184
