We seemed to not properly fallback to querying the database if the cache
did not contain a mailbox of a specific type. We may want to consider
caching mailboxes related to PIM data, but for now this is a quick fix.
Also fixes unit tests.
Bug: 5019661
Change-Id: Idcac0a6f15aa7f174890ae586f478fbb8f6e05b7
This will avoid cases where the user is permanently stuck, if we fail to
error out when a failed sync occurs or for any reason.
Bug: 5032763
Change-Id: Ib82a62add767d13dc844a90e5ea4147d3ada59dd
This info used to be shown in the top right of the status bar in GB but
that is no longer there. So before this change there was no way to tell
who you were sending as.
Bug: 5069124
Change-Id: I10d687a5fd221e65ecfc71b61cccf306492e27b8
Seems like POP envelopes could get into the database and tickle the
NotificationController even before we got the full subject/sender. Just
ignore those things until the basic info is loaded.
Bug: 5061271
Change-Id: Iadfbff8a1615d2644880f5cae3727768f4f9549a
- use title as ticker text
- also properly play notification sounds on new messages. Before, if you
left a notification unread, and a new message comes in, no sound would
be played. Since fixing that, it introduces another issue where on
initial sync, tons of new messages come in (and the sync could take > 1
min). We throttle them with a 15 second interval. The notification is
always updated to reflect the most up to date information, but sounds
will never be played closer than 15 seconds together.
Bug: 5020191
Bug: 5067059
Change-Id: I5ca474fd3b210ee856035bd78bd72931da80fe40
We were hiding the cc/bcc menu option, but on tablets that is a visible
button on the UI and not in the menu. whoops.
Bug: 5067057
Change-Id: I10ecaf115b589b9200529f6ea7703428456b8f0f
This prevents accessing a potentially closed cursor when doing batch
operations that will inevitable cause the the list to be reloaded (and
cursor to be invalidated) as the first of the messages are
touched
Bug: 5051730
Change-Id: I90328ee02eafe6ad238d8c57e88a3d96259f6547
When editing quick responses, displays an appropriate
message letting users know that there are no quick responses.
Additionally, the MessageCompose activity will only show "insert quick
responses" when at least one exists.
Bug: 4948180
Change-Id: I1191bd54d6e64f990b6d6830f0837579d7a6918c
Shortcut name is now the name of folder when it's a non-inbox
folder; otherwise it's the account name. Previously was
always account name.
Bug: 4903490
Change-Id: I15b4024d28db6a48464e423db887bdca99132423
If the account is in the process of being deleted, large parts of its
data may be being removed in parallel with notification processing. We
never handled this very well.
Bug: 5058333
Change-Id: Ic588b68438a9c5829e7e3fd5c8d52e21a037b9f4
- the container is the one that should be visible/gone - individual
fields should just be visible
- also do drive by javadoc fixes
Bug: 5052310
Change-Id: I45c4b6b7f94b2b62ac94eeff0ab97bf34c9e117a
We were hiding the search icon from the actionbar but didn't take into
account the use of a hardware button.
Bug: 5051691
Change-Id: I2e0bfd1d0bb4faa3238e567c02189894db5f6c44
This fixes a crash on account deletion, leaving around bad accounts that
were in limbo
Also remove a test for an unsupported operation
Bug: 5051951
Change-Id: Ieebc7f769075614ae1a656cf123d8ce0313e611d
On second thought, it's probably best to stick with the same account if
I can't find a particular folder. This at least is less jarring if the
user has multiple accounts.
Change-Id: Ifd5d631b220e260399681008ac17203f5451c8ff
- ensure that the commitAllowingStateLoss is actually not dangerous - we
only had one case I could see where we were actually doing a transaction
from a loader callback and I fixed that just now to not do any
transactions.
Change-Id: I21e11138f70eb2ce953a5ba54119ca46555f465d
* This CL fixes the referenced bug, but it does NOT explain how
mAccount; best guess is that the process was killed and then
restarted when the result from DPM was available.
* Assuming this is the case, we remove the background task loading
mAccount, avoiding a possible race.
* Also, it's not clear why clearNotifications didn't use the
account id argument; what if there's more than one account that
uses security? Filing a bug about this.
Bug: 5048912
Change-Id: I734834337ab6e409d77624e7c7370350de76becb
* Move AccountReconciler to the Email app (from EmailCommon)
* Ensure that Controller.deleteAccountSync() performs ALL actions
needed to clean up after an account deletion (delete attachment
files, reset policies, refresh the UI, etc.)
* Add reconcileAccounts() API to AccountService
* Remove accountDeleted() and restoreAccountsIfNeede() from the
AccountService API
* Remove unused callback
Bug: 4883073
Bug: 4767084
Change-Id: I43ffaf009db1a6f306bb0f2a74fb4dd3b2c4b966
