should be used.
From now on, UiUtilities.useTwoPane() should be used to see which UI
should be used.
You can pass the DEBUG_PANE_MODE extra when launching Welcome
to force which UI mode to use. (See the comment inside Welcome.)
Change-Id: Iefa3737e4979eb55f7986a9033ff9c6266d32f52
- Now we "uninstall" a fragment in Fragment.onDestroyView.
i.e. a fragment transaction is actually executed.
- Maintain our own "about to be removed" fragment list to avoid
double removal of a fragment.
Change-Id: I61328e0a09a7af00cbb0e6ba10a2d39c11b5c3dc
Previously, we were deciding whether to show the spinner if the returned cursor
had more than 1 element. However, the cursor would _always_ have more than one
element because it contains a header in addition to the account information.
We now save away the number of accounts and the size of the recent list into the
cursor so we can use those counts to determine if the spinner should be shown.
Both of these counts are necessary as we want to show the spinner even if there
is only a single account as long as it has at least one recent mailbox.
Change-Id: If428c496a548a25f6b5cd7301ddb5c0d6876750c
The ability to change mailboxes using the spinner is currently only implemented
for the two-pane UI. one-pane implementation will come in a future CL.
Change-Id: If72e9d9d607508553c918f5523e748e8a481ff84
This will make the message list a lot snappier.
We were using two different loaders for the message list; one is to load
the meta information and the other to load the actual message list.
Unify them using the CursorWithExtras pattern.
Change-Id: I02957bbca1b1fb74ca6eca14ad2535dfdbf03a5a
we need to return 5, post filtered, results. previously, we were returning
5, pre filtered, results. Also add a test to catch this condition.
Change-Id: Id25f4bf79081c42a2012e0e51b36142120c83b20
(this conflicted and the automerger situation was a disaster, so I
opted to skip immediately to help clear the backlog, now I'm
cherry-picking manually)
Using a TableLayout and large margins to shrink the input boxes causes
weird artifacts when you type past the size of the input box. Fixing the
padding is also not the right way, since stretching the input the whole
screen length is ugly. Just use a width.
Change-Id: I3009e7c66f729a96f404263d79e77d3ab4d831ae
Using a TableLayout and large margins to shrink the input boxes causes
weird artifacts when you type past the size of the input box. Fixing the
padding is also not the right way, since stretching the input the whole
screen length is ugly. Just use a width.
Bug: 4495679
Change-Id: I0e93fdd4f050b193338ca20cef6cba208547c9aa
- Now that fragment useage is simplified (e.g. no new fragment
creation for nested mailbox navigation), most of the fragment
operation code for 2-pane is reuseable for 1-pane as well,
so moeved it to the base class.
- Temporarily added "Show all folders" as a menu option on 1-pane.
- Added "opener account id/mailbox id" to the message view fragment.
They are not used by the fragment itself, but they're used
by the UI controller for the back navigation. (And now the UI
controller doesn't maintain the current IDs by itself; rather
it gets them from the currently-active fragment.)
- Use async fragment transaction on 1-pane too, now that it always
gets the current state from the active fragment.
- Changed the timing when we install fragments from onAttachFragment
to fragments' onActivityCreated. So now all installed fragments are
created.
TODO Now that all installed fragments are guaranteed to be created,
remove all special trealment for the fragment argument accessor.
(They were meant be safe to call before onCreate, but it's not
necessary any more.)
Change-Id: I0ed100c3f0b460835b164c0dc908ea483a4e46ee
Upon hitting reply all, all email addresses except that of the source message sender
go in the CC: field. Previously they all went in the TO: field. Updated 3 tests
in MessageComposeTests.java (testReplyAllAddresses1(), testReplyAllAddresses2(), and
testReplyAllAddresses3()) to reflect this new behavior.
Bug: 4534058
Change-Id: I852daebdd8843a45f685eecc67f757c87925bb6c
