Co-authored-by: Andrew Morgan <1342360+anoadragon453@users.noreply.github.com>

When a server leaves a room it may stop sharing a room with remote
users, and thus not get any updates to their device lists. So we need to
check for this case and delete those device lists from the cache.

We don't need to do this if we stop sharing a room because the remote
user leaves the room, because we track that case via looking at
membership changes.

If we detect that the remote users' keys may have changed then we should
attempt to resync against the remote server rather than using the
(potentially) stale local cache.

* Bump signedjson to 1.1

... so that we can use the type definitions

* Fix breakage caused by upgrade to signedjson 1.1

Thanks, @illicitonion...

MSC2260: change the default power level for m.room.aliases events

Fixes a bug where the alias would be added, but `PUT /directory/room/<alias>`
would return a 403.

I'm going to need another copy (hah!) of this.

We were sending device updates down both the federation stream and
device streams. This mean there was a race if the federation sender
worker processed the federation stream first, as when the sender checked
if there were new device updates the slaved ID generator hadn't been
updated with the new stream IDs and so returned nothing.

This situation is correctly handled by events/receipts/etc by not
sending updates down the federation stream and instead having the
federation sender worker listen on the other streams and poke the
transaction queues as appropriate.

Otherwise its just stale data, which may get deleted later anyway so
can't be relied on. It's also a bit of a shotgun if we're trying to get
the current state of a room we're not in.

We ended up only returning a single device, rather than all of them.

Implement updated auth rules from MSC2260

We just mark the fact that the cache may be stale in the database for
now.

These are easier to work with than the strings and we normally have one around.

This fixes `FederationHander._persist_auth_tree` which was passing a
RoomVersion object into event_auth.check instead of a string.

As using non-C locale can cause issues on upgrading OS.

Calling the invalidation function during initialisation of the data
stores introduces a circular dependency, causing Synapse to fail to
start.

This is so that we don't have to rely on pulling it out from `current_state_events` table.

Currently if a worker invalidates a cache it will be streamed to master, which then didn't forward those to other workers.

Clarifications to the workers documentation

Co-Authored-By: Brendan Abolivier <babolivier@matrix.org>

Signed-off-by: Jason Robinson <jasonr@matrix.org>

* Add note that user_dir requires disabling user dir
  updates from the main synapse process.
* Add note that federation_reader should have
  the federation listener resource.

Signed-off-by: Jason Robinson <jasonr@matrix.org>

There are quite a few places that we assume that a redaction event has a
corresponding `redacts` key, which is not always the case. So lets
cheekily make it so that event.redacts just returns None instead.

* a bit of debugging for media storage providers

* changelog