* Allow, but provide warnings via EditText.setError()
* Remove one last instance of password trim()
Bug: 2981433
Change-Id: I406a4f8b8f27cc5ce90424a8cafe88a677e72f45
Signed-off-by: Andy Stadler <stadler@google.com>
- Now tapping these To/Cc/Bcc/Subject labels moves the focus to
the corresponding edittext.
- Tapping the bottom part of the screen moves the focus to the
main EditText.
- Also use paddings instead of margins for the main EditText, to
expand the hitarea.
Bug 3366831
Bug 3367100
Change-Id: I9b5d18dcc9d7802bfcbd0160befcb008c784d9f7
* This fixes the case of:
* a device that does *not* support device encryption
* connecting to an account that *does* require device encryption
* but also supports "non-provisioned devices" (making the encryption
requirement optional.)
* Added unit test
Bug: 3367191
Change-Id: I894e68c4119a102dad02d2e0815fccdae1e87189
instead of showing all messages (e.g. messages in drafts, etc...) the "combined
inbox" view now only messages in the inbox folder. This is now identical to
the "combined inbox" view available in the full email UI.
bug 3368613
Change-Id: I0080b56cd2718a3dce82b279277c63c4f43e86dc
* Modified font colours for read/un-read emails
* Add chip colours when viewing mail from multiple accounts
* Add calendar icon if message has an invite attachment
* Update background of read/un-read emails
bug 3351761
Change-Id: Id59573d25a6988e9e869335f95778aad28b43912
Bug: 3301398 Phone->Android device in device admin setup screen
Bug: 3296594 email widget doesn't have a "tap to configure"
Bug: 2981433 should not trim leading/trailing spaces from passwords
Bug: 3360873 Attachment pre-caching should be a per-account setting
Bug: 3362752 Malformed message can prevent sync (error while decoding)
Change-Id: Ibed653987ccd5e0e1bd4e195f044c5c1cc06368f
- Use the fixed margin in portrait, as opposed to the fixed width we use
in landscape.
- Add 16 dip bottom margin for both landscape and portrait.
Bug 3335094
Change-Id: I082380048256aeef67673ed1b961af279fe2e702
After saving an attachment, we no longer automatically start the activity
associated with the attachment mime type. However, we still run the media
scanner to add supported media (e.g. music, pictures, etc...) to the media
content provider.
bug 3266378
Change-Id: I96985438316a33322437ff009fe7e9c597b1c70a
* Use getStorageEncryptionStatus() to check device status
* Also, check granted policy on USES_ENCRYPTED_STORAGE
Bug: 3346641
Change-Id: I9e9a45a6d1d3cf4714e27b69cdb5952c841c640d
We don't need to allow users to install applications directly from the email
client. Instead, application installation is a two step process; the user
must first save the APK and then find it on the filesystem.
If the user does not want to allow installation of applications from unknown
sources, we don't provide the ability to save.
NOTE: After saving, we still try to open the APK which generates an error
toast. We will be removing the auto-open-after-save feature in a separate CL.
bug 3351137
Change-Id: I0eb1bc8224a154792fe852757e4b23a3059f4392
When navigating away from a preferences screen and unsaved settings would be
lost, display a confirmation dialog. The user can either accept or cancel the
action. If canceling, the user is returned to the settings screen they were
currently on. Otherwise, they are taken to a new fragment (the exact
destination depends upon whether the user navigated "back" or selected another
header)
There is one additional change that needs to be made. In the case of navigating
to another header, we are notified _after_ the new header is selected. In this
scenario, the action is not cancelable and the user will lose any changes. We
must display an appropriate message when this happens. [note: this is the same
behaviour as when the user selects a breadcrumb]
bug 3327737
Change-Id: I4bd3b393a6323f3e63510e3ed08e4e1e745b04c4
The problem was that we tried to upgrade the db by re-creating a LocalStore,
without closing the existing connection. The db was still upgraded on the
filesystem, but the old connection kept looking at the old table definition
for whatever reason.
Also cleaned up checkAllTablesFound(). It now shows what columns are
missing when failed.
Bug 3349973
Change-Id: I80f4c67f1bd61cd5e1f12a632978b36a6e0a3b19
