| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
| |
left for a drop. This change fixes this while retaining the 'snap to vacant cell' ability added in Cupcake.
|
|
|
|
|
|
|
|
|
| |
When dragging between workspaces in rapid succession, the canceling of the
original long-press timer wasn't being triggered correctly. (When the timer
fires, it might read an invalid Workspace.allowLongPress() value.)
This patchset correctly cancels any pending long-press timers once a desktop
scroll begins, and we don't need to rely on the allowLongPress() value.
|
|\
| |
| |
| |
| | |
* changes:
Fixes #1596240. Optimize invalidate/draw passes by marking opaque views and avoiding drawing them. Whenever a View requests an invalidate its parent check whether the view is opaque or not. When the view is not opaque, the framework behaves as it used to. However, when a view is opaque, the parent marks itself as being dirty because of an opaque view. Its parent then does the same, and so on. When the framework then starts drawing the views, it does not draw views marked as dirty opaque. If a view is dirty opaque and receives an invalidate request from a non-opaque view, it then clears the dirty opaque flag and behaves as before.
|
| |
| |
| |
| | |
avoiding drawing them. Whenever a View requests an invalidate its parent check whether the view is opaque or not. When the view is not opaque, the framework behaves as it used to. However, when a view is opaque, the parent marks itself as being dirty because of an opaque view. Its parent then does the same, and so on. When the framework then starts drawing the views, it does not draw views marked as dirty opaque. If a view is dirty opaque and receives an invalidate request from a non-opaque view, it then clears the dirty opaque flag and behaves as before.
|
|\ \
| |/
|/|
| |
| |
| |
| | |
Merge commit 'c4c1f32159cf125f60c6057e5099aa3e0dc439cf' into donut
* commit 'c4c1f32159cf125f60c6057e5099aa3e0dc439cf':
AI 148592: Fix wallpaperchooser from previewing the default_wallpaper to wallpaper_lake. This issue is exposed
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
wallpaper_lake. This issue is exposed
when we override the default wallpaper.
BUG=1836838
Automated import of CL 148592
|
| |
| |
| |
| |
| |
| | |
BUG=1749373
Automated import of CL 143907
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
completes.
During the first boot upgrade, LauncherProvider will deleteHost() to clear out any old appWidgetId bindings. During the first boot, Launcher calls AppWidgetHost.startListening() to watch for widget updates. It also calls loadUserItems(), which loads data from LauncherProvider, triggering the database creation and deleteHost() call. Because deleteHost() removes any existing callbacks, any future widget updates are dropped on the floor. (This can currently be solved by rebooting, because there isn't an upgrade on subsequent boots.)
This bug was particularly evident on vfpioneer-userdebug builds, as there aren't any configuration changes that cause Launcher to be destroyed and recreated. (When destroyed and recreated, we startListening() again, and LauncherProvider doesn't call deleteHost().)
To handle this special case, Launcher creates a ContentObserver pointing at a specific URI, which the LauncherProvider notifies when the AppWidgetHost is reset through deleteHost(), allowing Launcher to correctly startListening() again.
BUG=1746104
Automated import of CL 143622
|
| |
| |
| |
| |
| |
| |
| |
| | |
first time, it spawns a loading thread for the workspace items. That loader is responsible for starting the drawer loader after it's done. Unfortunately, after a wipe data or a database upgrade, the worksapce loader could be interrupted by a ContentProvider notification change which would cause the loader to be cancelled and another workspace loader to be spawned. The new workspace loader, however, would not start the drawer loader because the appropriate flag was not set correctly. This change simply duplicates the flag from the old loader in the new one and all is well.
BUG=1725672
Automated import of CL 143294
|
| |
| |
| |
| |
| |
| |
| | |
Only happens when user changes orientation while editing, where getFolderById can return the wrong folder type for the given ID.
BUG=1740367
Automated import of CL 143057
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| | |
This was caused by the workspace optimization that draws only what is thought to be the current celllayout.
|
| | |
|
| |
| |
| |
| | |
leaves Home by launching an application.
|
|\ \ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
- all public apis and framework changes have been reviewed by relevant folks in our branch (e.g romainguy)
- all new public apis are @hidden; they will still get reviewed by api council once we're in git
- other than that, it's mostly GlobalSearch and search dialog stuff, a new apps provider, and some tweaks
to the contacts provider that was reviewed by jham
Automated import of CL 147564
|
|/ /
| |
| |
| |
| |
| |
| |
| | |
This change adds nearby-area searching when dragging desktop items into already-occupied cells. This approach tries harder to find a matching area, instead of strictly rejecting invalid moves.
We also draw a "snag" during the drag to show where an item would be dropped, but only if we would drop into an alternative cell. This gives users better feedback about where things will drop.
http://b/issue?id=1634887
|
| |
| |
| |
| |
| |
| |
| | |
Original author: jsharkey
Merged from: //branches/cupcake/...
Automated import of CL 143908
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
first boot completes.
During the first boot upgrade, LauncherProvider will deleteHost() to clear out any old appWidgetId bindings. During the first boot, Launcher calls AppWidgetHost.startListening() to watch for widget updates. It also calls loadUserItems(), which loads data from LauncherProvider, triggering the database creation and deleteHost() call. Because deleteHost() removes any existing callbacks, any future widget updates are dropped on the floor. (This can currently be solved by rebooting, because there isn't an upgrade on subsequent boots.)
This bug was particularly evident on vfpioneer-userdebug builds, as there aren't any configuration changes that cause Launcher to be destroyed and recreated. (When destroyed and recreated, we startListening() again, and LauncherProvider doesn't call deleteHost().)
To handle this special case, Launcher creates a ContentObserver pointing at a specific URI, which the LauncherProvider notifies when the AppWidgetHost is reset through deleteHost(), allowing Launcher to correctly startListening() again.
Original author: jsharkey
Merged from: //branches/cupcake/...
Automated import of CL 143776
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
starts for the first time, it spawns a loading thread for the workspace items. That loader is responsible for starting the drawer loader after it's done. Unfortunately, after a wipe data or a database upgrade, the worksapce loader could be interrupted by a ContentProvider notification change which would cause the loader to be cancelled and another workspace loader to be spawned. The new workspace loader, however, would not start the drawer loader because the appropriate flag was not set correctly. This change simply duplicates the flag from the old loader in the new one and all is well.
Original author: romainguy
Merged from: //branches/cupcake/...
Automated import of CL 143295
|
| |
| |
| |
| |
| |
| |
| |
| | |
Only happens when user changes orientation while editing, where getFolderById can return the wrong folder type for the given ID.
Original author: jsharkey
Merged from: //branches/cupcake/...
Automated import of CL 143232
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
to avoid the dreaded "blank drawer" bug.
Original author: romainguy
Merged from: //branches/cupcake/...
Automated import of CL 143187
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
|/ |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|