* Some reasons this needs to exist for it to run on my machine? I don't think it would hurt to have it in other machines so
* Add options to member model
* Add Privacy to member embed
* Added member privacy display list
* Update database settings
* apparetnly this is nolonger needed?
* Fix sql call
* Fix more sql errors
* Added in settings control
* Add all subject to system privacy
* Basic API Privacy
* Name privacy in logs
* update todo
* remove CheckReadMemberPermission
* Added name privacy to log embed
* update todo
* Update todo
* Update api to handle privacy
* update todo
* Update systemlist full to respect privacy (as well as system list)
* include colour as option for member privacy subject
* move todo file (why was it there?)
* Update TODO.md
* Update TODO.md
* Update TODO.md
* Deleted to create pr
* Update command usage and add to the command tree
* Make api respect created privacy
* Add editing privacy through the api
* Fix pronoun privacy field in api
* Fix info leak of display name in api
* deprecate privacy field in api
* Deprecate privacy diffrently
* Update API
* Update documentation
* Update documentation
* Remove comment in yml
* Update userguide
* Update migration (fix typo in 5.sql too)
* Sanatize names
* some full stops
* Fix after merge
* update migration
* update schema version
* update edit command
* update privacy filter
* fix a dumb mistake
* clarify on what name privacy does
* make it easier on someone else
* Update docs
* Comment out unused code
* Add aliases for `member privacy all public` and `member privacy all private`
- DbConnectionFactory renamed to "Database", will now be the primary entry point for DB stuff
- Created IPKConnection interface mostly containing async extensions to IDbConnection, use this going forward
- Reworked the Connection/Command wrappers (that have performance/logging extensions)
- Probably more stuff that I forgot???
- Allowed adding ephemeral(ish) views and functions
- Moved message_count to a concrete database field
- Moved most proxy logic to a stored procedure
- Moved database files around and refactored schema manager
- Move reaction handlers to the ReactionAdded event instead of
ProxyService
- Split tag matching off into ProxyTagParser
- Split autoproxy matching off into Autoproxier
- General cleanup and simplification
This happened because we cache Task instances, not the values, and a failed task is still a *valid* task that can repeatedly be awaited. This lead to the Task being saved even if it failed, and the same exception constantly being re-thrown.
This fix invalidates the cache if it finds a failed Task, and lets it retry fetching as normal (hopefully successfully this time).
