GitLab

Change-Id: Ifdea363c8ee351a42c630ce84cf6f8dff7ee7e7c

Change-Id: I90d8c69ac30612ace077df0b48c44b7628763ad5

Change-Id: Ie3b00ac372907ea32464ab334ee22f9c03647b9a

Change-Id: Icdc77662827d10bd0f8ab32616d380db29bd94bb

Change-Id: I7d49b84bb61ca354fef309367d76d79deaa2e83b

Bug: 2983765
Change-Id: I1fdd33ea8ee4fbc79a75ce0450f04d8e18915975

The pause is mostly notable when a sync is happening at the same time.

The related function addItemToDatabase() is unchanged by this CL
though as it populates item.id and would need more invasive changes
elsewhere.

Change-Id: I6538445aeb15d9efedf99f262c39f6175f7f1e0e

Bug: 3066123
Change-Id: Icbdf3ef241ec44b5569e7490c1b950d80699ddc8

Change-Id: I5505d267076de39cb3fabd0fa7419adc3f619d39

Bug: http://b/issue?id=3053235

This is a proper way to do what was done in https://android-git.corp.google.com/g/#change,71522

The longAxisEndPadding used to be equal to button_bar_height.
Increased it to match proposed mock from designers and increase space
between last row and hot seats buttons.

Change-Id: I31dfba0a3e90a30f18d42b16681592d9527517d4

The problem was that we were doing startLoader, but once the
apps are loaded, all that does is re-bind everything.  Before
a recent change of mine, we were clearing mWorkspaceLoaded and
mAllAppsLoaded.  But that's not good either, because it meant
that there needed to be a lock that the main thread and the
worker thread were both acquiring, which was causing ANRs.

Instead, now we schedule two messages - one to add the apps
back into the list, and one to update everything to use
those new icons.

Bug: 3065061
Change-Id: If9d36d4b55a7ba72fcf48a5ed8941bb9466bebe6

This reverts commit 372ca973.

Change-Id: I5c5df72067bd17b14c8afc7b7869a50a0198f9ac

Change-Id: Id10c202123114a32708d7def3942b900f1d5933b

Change-Id: I95ebf95f6eb0d24a88e1726acbb9e2498965d180

Strange behavior: cell height has to be increased to increase padding.
I guess this is because the cells have to fit in the screen height and are
rescaled, with a top alignement thus creating padding at the bottom.

Bug: http://b/issue?id=3053235

Change-Id: I727ab091a736916b7efc1c05a80c5b006bc56f67

Bug:3049178 Launching the phone app should always take the user to the dialer
Change-Id: I61edffa37a211e205e8b2192e6a81963eff54ee7

Change-Id: I9703f80b34d05b2cfa6239652eae36d1ebcde6e5

We could do this more efficiently by only updating the packages that
have changed, but that would involve a big rewrite.

Bug: 3020854
Change-Id: Ie7e3bee20f45dc5efa93d960902e9dfb79c4a3a5

Merge "Add a script to pull launcher.db off the device and pretty print an html file with the contents." into gingerbread

Change-Id: I4640fad71e5f77e0dbcbe4df4204bf36a03d3562

Change-Id: Ic9f90a30833bfafb7d76dcc851a92892b9a2a9b7

Change-Id: Iadedd45343b749561216f7b2d9ef4c3de0d4034c

Change-Id: If1304e1417cb8ba4dee8657cd619718abba277f0

Change-Id: I22c1e4a8b932fa48af57e2eecedb6ac4fe6c85fa

Change-Id: I56d726080968777eb5ccdc8826a872977cf7868c

It's possible that the app doesn't implement the component as promised, and the launcher shouldn't
crash.

Bug: 3005141
Change-Id: Ieac51433a467fd12160247bfd281828a69f5f7d2

Change-Id: Ie0dd3e3ea930de8676d4e085721cab5844386f66

Fixes a bunch of StrictMode violations and removes TODOs.

Change-Id: I886bd071d4a8336c228d29c517a8ffc5223c0623

Logging indicated a 1.2 second improvement in Launcher start time
during boot (when single-threaded yaffs is very heavily contended).

BUG=2947474

Change-Id: Ie72cf02c5eacb27c6d874b6d67a0bb2de7f6b74a

You still can't modify the contents of the workspace while it's loading,
but this makes it appear a little bit more responsive.  Launching apps
while the launcher is loading has always been possible.

Bug: 2748595
Change-Id: Icf252beaee5b801b30216202ce6032c4d565cde3