-
Notifications
You must be signed in to change notification settings - Fork 269
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Reconcile or unify RoomEvents
event chunks with AllEventsCache
events
#3886
Comments
After we've implemented proper persistent storage for events, I'd expect that the second |
For what it's worth, we might even not need migrations for the sqlite backend, if we consider using sqlite's |
I know it's not something easy, but I think we may want to bench having a new column for the event type vs. using On the paper, it's not particularly elegant to use |
Unfortunately, this is not sufficient: the But we have a tangential problem, that redacted events are still stored in clear in the database, and they shouldn't be. Otherwise, a user may see the content of a redacted event, if they're tech savvy enough to find the sqlite database. Probably, an event that the event cache is aware of and then that is redacted should be replaced in the database by the redacted for. We can keep the redaction event around too, in case we want to display the reason later. This would avoid the special-casing for the |
We actually can't use |
In
EventCache::save_event
there is this comment that says:And we have 2 separate caches now: one for chunks of events for timelines of rooms (
RoomEvents
), which come from room timeline updates during syncs and back pagination, and another one for any event in any room and their related events (AllEventsCache
), which may come from any source.I believe the end goal here is using a persistent storage to hold a cache that unifies both of these, and it's probably related to the 'implement a basic reconciliation of events; don't handle separate connected components of events yet' point in #3058. I'm opening this issue to both discuss if this is the way we want to proceed and, if it is, keep some recording of it.
The text was updated successfully, but these errors were encountered: