new timeout was set that was smaller than the previously set timeout
> Using uptimeMillis in Alarm to avoid errors due to system time
changes
> Adding an extra check in Wait in case Thread.sleep eats up
all the timeout
Change-Id: Id1fac5e8fdb81a0c3c7a6a5e50586b2a2f180d06
- We take the first 4 in sorted order, except we remove up to 2
static shortcuts to make room for dynamic ones if they exist.
- Added ShortcutFilterTest with testSortAndFilterShortcuts().
This asserts that the filtered list is sorted and has the
expected number of static and dynamic shortcuts based on
inputs with various amounts of each.
Bug: 28980830
Change-Id: I832b6b21144f17c74bb8b90a840d6620e99911b8
provider. This allows OEMs to keep the user's homescreen intact while
changing the default home app package.
Bug: 28536314
Change-Id: Ibebfd7dd33aa2cbd9ca28d2d611dd0a4a5971444
> Renaming it to simply DropTargetBar
> Moving AppInfo to the top bar as well
> The workspace pages will extend to the top edge (minus some padding).
Since the QSB is no longer displayed on top of every page, there is
no reason to reserve the space.
> In spring-loaded mode, the workspace cell layout will scale enough
to make room for the drop target bar at the top
Change-Id: I2baf607310335dd576c9d9fcbb75ab708f47ac03
The QSB will only be resent on the first screen of the workspace
covering the full width of the first row. If will not be movable.
The first screen of the workspace will not be movable.
The searchDropTargetBar no longer contains the QSB (it can be
renamed in aseparate cl).
Refactoring all QSB related logic by moving it to a custom view
inflated only using xml.
Change-Id: Icb4fd6eb855df1af15f685961c38351bf4fd4f4a
The logs are kept for at max 48 hours. It uses two log files and switches
between the two based on the day of the year.
Change-Id: I9a99499b3445a62f29f62a5cd13db20b1783bcd3
Rational/Background:
@MediumTest: An integration or functional test that validates the
interaction of one or more application modules on a single machine.
(should run under 1 sec: Usually get killed after 5 min)
@LargeTest: A system test, integration test, or end-to-end test that
verifies that a complete application works and accounts for the behavior
of external subsystems.
(Usually get killed after 15 min)
Change-Id: I8b668af3d2ca09c43ad45c6fd72b1d0d80c6b7d3
- Missing a comma in one of the focus logic test matrices
- Disable QuickAddwidgetTest, because it isn't that useful and is
flaky by nature.
Change-Id: Ia85fa578ea563ca525419b68ffcf90cfef8f86d0
The grid is migrated in steps where each step consists of at max one column change and at max one row change.
Adding some unit tests for GridMigrationLogic
Bug: 25958224
Change-Id: Ie54e872ea0925cc4c463edbba0a7201d62b373a0
This would have caught the race condition bugs that caused the "specified child
already has a parent" IllegalStateException (bug 23896857).
Change-Id: I3b408d21113237b0c89a0a7161f504596212bd58
This should be the last keyboard CL.
- Fix bug: couldn't focus All Apps button in some cases when the All
Apps column was skipped over. Also added test case for this.
- Stop explicitly passing countX and countY to handleKeyEvent, as
these had to match the matrix dimensions anyways.
- Rename createSparseMatrix() - there were 3 methods of the same name,
but all had different purposes. This is confusing both from a
readability standpoint and also when looking at stack traces.
Change-Id: I08ba8411674fcea43a608856c114dee8dbd22398
- Handle NextPageFirstItem as first focusable item in reading order
- Handle PreviousPageLastItem as last focusable item in reading order
- Check the hotseat after the workspace in both cases above
- Dpad horizontal navigation (left/right) uses these as a last
resort (Rule3) to guarantee an item takes focus if a page exists
Note that it is necessary to search for a focusable item because
widgets are not yet focusable.
Bug: 25591057
Change-Id: I953648bd76c657d660a38427fdd4108bf9963c23
The All Apps button creates a number of edge cases, mainly because it
causes the hotseat to sometimes have an extra column than the workspace.
Previously, we sort of swept these problems under the rug by simply
ignoring the All Apps button if other icons were present in the hotseat,
with the assumption that those other icons should get focus instead of
the All Apps button. (If possible, we want to stay in the same column
when moving from the workspace to the hotseat.) But this doesn't always
work, as in the attached bug where the hotseat doesn't get focus at all
when the All Apps button is an obvious candidate for it.
By adding a specialized column in the focus matrix for the All Apps
button, we ensure that moving down to the hotseat stays within the
original column when possible, while also allowing the focus to switch
to the All Apps button if appropriate. Furthermore, we take care to skip
over the All Apps column when necessary in order to maintain all
previous functionality.
Bug: 25590522
Change-Id: I5d6a8ee69de8834314c4689246fe7d54329b2eef
- Seperate search bar height from its top and bottom paddings
- Report this height plus 8dps top and bottom for the widget options
- Use a separate bottom padding to adjust spacing between QSB and
the workspace
- Add tests to verify the height of the search bar in both normal and
tall modes (ran on Nexus 6, 7, and 10)
Change-Id: Ia7557785d1b38194eee869d0460456e8f33447d5
b/19381790
b/16351792
TL;DR;; Previously, when RIGHT is handled on the right most column of the
current page or when LEFT is handled on the left most column, the next
icon of focus is next page 'first' icon or the previous page 'last icon'.
With this change, the row information is preserved when trying to locate
an icon to give focus in the next/previous page.
Next CL: long awaited unit tests that capture corner cases for different
orientation/ device configuration.
Change-Id: I5278bed45275b3e4cb39fb698df35f90bb45a415
