Previous CL https://android-git.corp.google.com/g/55138 was submitted
before the suggested change was uploaded.
setSelection() was called twice by onTouchEvent, once to the previous
position, and then to the new position (unless the IME got displayed,
so that in that case the cursor is not moved).
The second call was actually triggering a call for a scroll of 0 since
the view is already displaying the cursor. This scroll is filtered out by a
shortcut in ScrollView. The first setSelection's scroll does not have
the same issue (since the previous cursor's position is out of screen
and requires a scroll) and it is then applied, effectively moving the cursor
to a new position but scrolling to the previous position.
The fix is to call setSelection only once, after the IME has been asked
to display. The cursor is moved to the old/new position depending on the
resultCode in onReceiveResult in CommitSelectionReceiver.
Bug: http://b/issue?id=2778954
Change-Id: I157d7ae451574e8a6fffb894eb1d6880acfe01d1
Add device class constant for PROFILE_HID and check for
Device.Major.PERIPHERAL when checking for class matches.
Change UUID for HID support.
Fix bug in BluetoothService.
Change-Id: Ie5361257d90fd749249e915824564bc3d79fb95d
setSelection() was called twice by onTouchEvent, once to the previous
position, and then to the new position (unless the IME got displayed,
so that in that case the cursor is not moved).
The second call was actually triggering a call for a scroll of 0 since
view is already displaying the cursor. This scroll is filtered out by a
shortcut in ScrollView. The first setSelection's scroll does not have
the same issue (since the previous cursor's position is out of screen
and requires a scroll) and it then applied, effectively moving the cursor
to a new position but the scrolling to the previous position.
The fix is to call setSelection only once, after the IME has been asked
to display. The cursor is moved to the old/new position depending on the
resultCode in onReceiveResult in CommitSelectionReceiver.
Bug: http://b/issue?id=2778954
Change-Id: I0983fdf18993e63b230e093e703f95efe6e2d7a8
Merge commit 'ef8f96a717554a85aa63a60f8c61be2afbd12a5c' into gingerbread-plus-aosp
* commit 'ef8f96a717554a85aa63a60f8c61be2afbd12a5c':
Start of work on passing around StrictMode policy over Binder calls.
Merge commit '28333315c6f62b303de2ec0ac9e48158ed419f06' into gingerbread-plus-aosp
* commit '28333315c6f62b303de2ec0ac9e48158ed419f06':
Added support for the GL_TEXTURE_EXTERNAL target
Added ANRs handling.
Added event injection.
Fixed a NPE ActivityManagerServer writing ANRs to the drop box.
Fixed HOME key interception.
Fixed trackball reporting.
Fixed pointer rotation in landscape mode.
Change-Id: I50340f559f22899ab924e220a78119ffc79469b7
This is (intendend to be) a no-op change.
At this stage, Binder RPCs just have an additional uint32 passed around
in the header, right before the interface name. But nothing is actually
done with them yet. That value should right now always be 0.
This now boots and seems to work.
Change-Id: I135b7c84f07575e6b9717fef2424d301a450df7b
Merge commit '9afeda0519c4f4892611565d678f01d269863bb8'
* commit '9afeda0519c4f4892611565d678f01d269863bb8':
Manual merge of change 53051 from master (DO NOT MERGE)
Merge commit 'e47e3f3855a062ba0338a57eeda2f12a0f7a1fa8' into gingerbread-plus-aosp
* commit 'e47e3f3855a062ba0338a57eeda2f12a0f7a1fa8':
Manual merge of change 53051 from master (DO NOT MERGE)
