Not quite sure why/when this happens - every stack trace is not our
code, but I do get an ClassNotFoundException for Notification$Type with
the new reverted code.
The notification fetching (worker) then stops/crashes so I never get an
Android notification.
It might have something to do with
https://github.com/tuskyapp/Tusky/pull/3732 ?
dc9e9f2aeb
modifed the code that fetched the value of EXTRA_NOTIFICATION_TYPE in an
intent, to use getSerializable().
However, the value was being placed in to the intent using putString().
This caused an exception when trying to update the summary notification,
so it would never update.
```
java.lang.ClassCastException: java.lang.String cannot be cast to com.keylesspalace.tusky.entity.Notification$Type
at com.keylesspalace.tusky.components.notifications.NotificationHelper.updateSummaryNotifications(NotificationHelper.java:321)
at com.keylesspalace.tusky.components.notifications.NotificationFetcher.fetchAndShow(NotificationFetcher.kt:87)
at com.keylesspalace.tusky.components.notifications.NotificationFetcher$fetchAndShow$1.invokeSuspend(Unknown Source:14)
```
Fix this by placing the value in to the intent using putSerializable(),
to match how it will be fetched.
Before, intent creation was scattered across multiple sites, with account switching logic in both `ComposeActivity` and `MainActivity`.
Now, intents are only created in `MainActivity` Companion, and account switching only occurs in `MainActivity`
Fixes#3695
Update to Kotlin 1.9.0 and migrate to newer language idioms.
- Remove unnecessary @OptIn for features migrated to mainstream
- Use `data object` where appropriate
- Use new enum `entries` property
Fix a crash where workers, in some conditions, should show a notification. These are sent to a dedicated channel with no importance.
Convert NotificationWorker to a CoroutineWorker and remove its use of `runBlocking`.
Fixes#3754
- Extend what was `NotificationWorkerFactory` to `WorkerFactory`. This
can construct arbitrary Workers as long as they provide their own
Factory for construction.
The per-Worker factory contains any injected components just for that
worker type, keeping `WorkerFactory` clean.
- Move `NotificationWorkerFactory` to the new model.
- Implement `PruneCacheWorker`, and remove the code from
`CachedTimelineViewModel`.
- Create the periodic worker in `TuskyApplication`, ensuring that the
database is only pruned when the device is idle.
Don't use `accountManager.activeAccount` throughout the code.
Instead, get the active account once, and use that over the life of the viewmodel.
As shown in https://github.com/tuskyapp/Tusky/issues/3689#issuecomment-1567219338 the active account can change before the view model is destroyed, and if that happens account information for the account will be written to the wrong account.
When fetching:
- Maintain a marker with the position of the newest fetched notification
- Use the marker to determine which notifications to fetch
- Fetch notifications with min_id to ensure that none are lost
- Update the marker as necessary
- Perform a one-time immediate fetch of notifications on startup
When creating notifications:
- Identify each notification with tag=${MastodonNotificationId}, id=${account.id}
- Remove activeNotifications field, it's no longer necessary
- Use the tag/id tuple to reliably identify existing notifications and avoid creating duplicates
- Cancelling notifications for an account must iterate over all the notifications, and individually remove the notifications that exist for that account.
- Limit notifications to a maximum of 40 (excluding summary notifications)
- Remove notifications (oldest first) to get under this limit
- Rate limit notification creation to 1 per second, so the OS won't drop them
Adjust the summary notification:
- Ensure the summary notification and the child notifications have the same group key
- Dismiss the summary notification if there is only one child notification
NotificationClearBroadcastReceiver is no longer needed, so remove it, and the need for deletePendingIntent.
Fixes#3625, #3539
* Fix saving changes to statuses when editing
With the previous code backing out of a status editing operation where changes
had been made (whether it was editing an existing status, a scheduled status,
or a draft) would prompt the user to save the changes as a new draft.
See https://github.com/tuskyapp/Tusky/issues/2704 and
https://github.com/tuskyapp/Tusky/issues/2705 for more detail.
The fix:
- Create an enum to represent the four different kinds of edits that can
happen
- Editing a new status (i.e., composing it for the first time)
- Editing a posted status
- Editing a draft
- Editing a scheduled status
- Store this in ComposeOptions, and set it appropriately everywhere
ComposeOptions is created.
- Check the edit kind when backing out of ComposeActivity, and use this to
show one of three different dialogs as appropriate so the user can:
- Save as new draft or discard changes
- Continue editing or discard changes
- Update existing draft or discard changes
Also fix ComposeViewModel.didChange(), which erroneously reported false if the
old text started with the new text (e.g., if the old text was "hello, world"
and it was edited to "hello", didChange() would not consider that to be a
change).
Fixes https://github.com/tuskyapp/Tusky/issues/2704,
https://github.com/tuskyapp/Tusky/issues/2705
* Use orEmpty extension function
* Add UI for selecting post language
* Apply selected language when sending status
* Save/restore post language with drafts
* Fall back to english if the configured language isn't found in the locale list (no-NB)
* Remove comment about no_NB
* Move language dropdown to top of compose view
* Preserve language when redrafting
* Set default language to target post's language when replying
* Add Tusky license header to new source file
* Tweak language dropdown button width
Fixes#793.
This is an implementation for push notifications based on UnifiedPush
for Tusky. No push gateway (other than UP itself) is needed, since
UnifiedPush is simple enough such that it can act as a catch-all
endpoint for WebPush messages. When a UnifiedPush distributor is present
on-device, we will by default register Tusky as a receiver; if no
UnifiedPush distributor is available, then pull notifications are used
as a fallback mechanism.
Because WebPush messages are encrypted, and Mastodon does not send the
keys and IV needed for decryption in the request body, for now the push
handler simply acts as a trigger for the pre-existing NotificationWorker
which is also used for pull notifications. Nevertheless, I have
implemented proper key generation and storage, just in case we would
like to implement full decryption support in the future when Mastodon
upgrades to the latest WebPush encryption scheme that includes all
information in the request body.
For users with existing accounts, push notifications will not be enabled
until all of the accounts have been re-logged in to grant the new push
OAuth scope. A small prompt will be shown (until dismissed) as a
Snackbar to explain to the user about this, and an option is added in
Account Preferences to facilitate re-login without deleting local drafts
and cache.
* Add notification bell button, API endpoints and new relationship field
* Add notification type for subscriptions
* Add subscriptions to legacy notification filtering
* Update schemas
* Fix build
* Make rewrite static method into method of Notification class, fix getNotificationText
* Mastodon wording for subscriptions
* Perform manual isolation of display names etc. instead of relying on BidiFormatter.
Fixes#1921
* Make follow request notification header formatting more like other notifications
UlrichKu
Levi Bard
Konrad Pozniak
Nik Clayton
Goooler
Eva Tatarka
Peter Cai
kyori19
Ivan Kupalov
Alibek Omarov