If the verify_request.deferred has already completed, then `remove_deferreds`
will be called immediately. It therefore might resolve the server_to_deferred
deferred while there are still other requests for that server in flight.

To avoid that, we should build the complete list of requests, and *then* add the
callbacks.

Define that it is run with no log context, and make sure that happens.

If we aren't careful to reset the logcontext, we can't bung the deferreds into
defer.gatherResults etc. We don't actually do that directly, but we *do*
resolve other deferreds from affected callbacks (notably the server_to_deferred
map in _start_key_lookups), and those *do* get passed into
defer.gatherResults. It turns out that this way ends up being least confusing.

... to make it easier to see what's going on.

This is a precursor to factoring some of this code out.

There's no need for this to be a nested definition; pulling it out not only
makes it more efficient, but makes it easier to check that it's not accessing
any local variables it shouldn't be.

This might make the cache slightly more efficient.

preserve_fn is a no-op unless the wrapped function returns a
Deferred. verify_json_objects_for_server returns a list, so this is doing
nothing.

Fix a bug where we could end up firing off multiple requests for server_keys
for the same server at the same time.

- allows sysadmins the ability to lock down their servers so that people can't
send their users room invites.

Demonstration of how you might add some hooks to filter out spammy events.

We should only kick guest users if the guest access event is authorised.

I'm still unclear on what the intended behaviour for
`verify_json_objects_for_server` is, but at least I now understand the
behaviour of most of the things it calls...

Don't filter out current state events from timeline

Add support for event_id_only push format

Correctly handle leaving room in /key/changes

as really it's part of the event ID

Param in the data dict of a pusher that tells an HTTP pusher to
send just the event_id of the event it's notifying about and the
notification counts. For clients that want to go & fetch the body
of the event themselves anyway.

Send down device list change notif when member leaves/rejoins room

do tox install with pip -e

PR #2413 added an issue template, but just adding files to the project
directory upsets the packaging scripts: we need to explicitly include or
exclude them.

Move the template into a .github directory to make that easy, and to de-clutter
the root a bit.

- this ensures we end up with a working virtualenv which we can use for other
things.

Document known to work postgres version

.. because I want to make the 'install_and_run' script useful for non-synapse
jobs, which do not accept --python. In any case we set up the path here, so
sytest shouldn't be guessing it.