On rooms with the `trusted` access level, any authenticated user
connected to the meeting can now manage the lobby. Requested by
several organizations, and a step toward generalized lobby
management once hubs and groups land (same organization only).
Being authenticated is not enough to grant the capability: a
`trusted` room means "trusted to join", not "trusted to decide who
else joins from outside the call". The new `CanManageLobby`
permission therefore also requires the requester to be currently
connected to the meeting, verified against LiveKit and failing
closed, like `IsPresentInMeeting`. The access level itself is never
cached and always read fresh, so an owner switching the room back to
`restricted` revokes the capability on the very next request - the
one guarantee we did not want to trade for performance.
Performance is traded elsewhere: the waiting list is polled by every
lobby manager, and on a trusted room that audience grows from a few
admins to potentially the whole meeting. Hitting LiveKit once per
poll per participant would not survive that fan-out, so presence is
memoized in Redis (`PresenceCache`, `PRESENCE_CACHE_TIMEOUT`, 1h).
Entries are created lazily because only the minority of participants
who actually manage a lobby ever need one, and only positive answers
are cached because a sticky negative would lock out someone joining
right after a miss for the whole TTL. Eager invalidation on
`participant_left`, `room_finished` and admin kick keeps the cache
honest; the TTL is the safety net when an event is lost, and its
value bounds how long a departed participant could still act.
Trade-offs in this v0:
* `PRESENCE_CLEAR_ON_PARTICIPANT_LEFT` gates the eager invalidation
on `participant_left`: its cost is one Redis DELETE per departure,
for every departure, so we want to be able to measure it in
production and turn it off independently of the feature. When
disabled, invalidation relies on `room_finished` and the TTL only,
widening the stale window above.
* This can put non-trivial pressure on the cache at scale; the
rollout will need to be monitored closely.
* The `participant_left` webhook must be enabled in the LiveKit
deployment, otherwise eager invalidation silently degrades to the
TTL-only behavior.
Introduce a new permission class that verifies the caller making a
request is both authenticated and actually present in the call.
It will be used to gate actions that require the user to be live in
the room, for example:
* allowing someone in from the waiting room
* promoting another participant to a different role
More generally, this covers every action where, for security
reasons, we need to make sure the user is truly present in the call
and that someone is not reusing their cookie as an API key.
Enable any participant to mute others when the room configuration
allows it. This is enabled by default for all meetings unless
explicitly disabled by an administrator.
Privileged users retain the ability to mute any participant
regardless of the room configuration.
Add FILE_UPLOAD_ENABLED setting (default to False, to avoid
a breaking change). Also adds a max_count_by_user sub setting
to restrict the number of uploaded files per user.
For the coming features we will need to store files on the meet side.
(for instance user backgrounds).
This commits adds a new Model to manage files, and the associated
serializers & viewsets. All are tested.
This work was heavily inspired by the work done by our friends at
https://github.com/suitenumerique/drive
It build on the same architecture design (upload directly to S3 but
download goes through our proxy), but model is much much simplier
(no folders, no file sharing, etc.).
Refactor feature flag mechanism from Django permission classes to custom
decorator that returns 404 Not Found when features are disabled instead
of exposing API structure through permission errors.
Improves security by preventing information disclosure about disabled
features and provides more appropriate response semantics. Custom
decorator approach is better suited for feature toggling than Django's
permission system which is designed for authorization.
Allow any user, anonymous or authenticated, to start subtitling
in a room only if they are an active participant of it.
Subtitling a room consists of starting the multi-user transcriber agent.
This agent forwards all participants' audio to an STT server and returns
transcription segments for any active voice to the room.
User roles in the backend room system cannot be used
to determine subtitle permissions.
The transcriber agent can be triggered multiple times but will only join a
room once. Unicity is managed by the agent itself.
Any user with a valid LiveKit token can initiate subtitles. Feature flag
logic is implemented on the frontend. The frontend ensures the "start
subtitle" action is only available to users who should see it. The backend
does not enforce feature flags in this version.
Authentication in our system does not imply access to a room. The only
valid proof of access is the LiveKit API token issued by the backend.
Security consideration: A LiveKit API token is valid for 6 hours and
cannot be revoked at the end of a meeting. It is important to verify
that the token was issued for the correct room.
Calls to the agent dispatch endpoint must be server-initiated. The backend
proxies these calls, as clients cannot securely contact the agent dispatch
endpoint directly (per LiveKit documentation).
Room ID is passed as a query parameter. There is currently no validation
ensuring that the room exists prior to agent dispatch.
TODO: implement validation or error handling for non-existent rooms.
The backend does not forward LiveKit tokens to the agent. Default API
rate limiting is applied to prevent abuse.
Generalize error message in HasPrivilegesOnRoom permission class to reflect
its broader usage beyond just recording contexts. Improves clarity when
this permission check fails in various application scenarios.
I've protected this endpoint with a feature flag, and an authentication
class, as it will be exposed on the public internet.
I've tried to keep the viewset logic as minimal as possible, I've
to ship smth and will continue iterating on this piece of code.
At some point, abstracting webhook endpoint and authentication class
might be beneficial for the project. YAGNI as of today.
The LiveKit egress worker interactions are proxied through the backend for
security reasons. Allowing clients to directly use tokens with sufficient
grants to start recordings could lead to misuse, enabling users to spam the
egress worker API and potentially initiate a DDOS attack on the egress
service. To prevent this, only users with room-specific privileges can
initiate recordings.
We make sure only one recording at the time can be made on a room.
The requested recording mode is stored so it can be referenced later when
the recording is saved, triggering a callback action as needed.
A feature flag was also introduced for this capability; while this is a simple
approach, a more robust system for managing feature flags could be valuable
long-term. For now, KISS (Keep It Simple, Stupid) applies.
The viewset endpoints were designed to be as straightforward as possible—
let me know if anything can be improved.
Implements routes to manage recordings within rooms, following the patterns
established in Impress. The viewset exposes targeted endpoints rather than
full CRUD operations, with recordings being created (soon) through
room-specific routes (e.g. room/123/start-recording).
The implementation draws from @sampaccoud's initial work and advices.
Review focus areas:
- Permission implementation choices
- Serializer design and structure
Credit: Initial work by @sampaccoud
Recent updates of dev/ruff and dev/pylint dependencies led
to new linting warnings.
Pylint 3.2.0 introduced a new check `possibly-used-before-assignment`,
which ensures variables are defined regardless of conditional statements.
Some if/else branches were missing defaults. These have been fixed.
I have updated all references of "Impress" to "Meet".
Migrations were manually updated and not regenerated. Never-mind,
they all will be squashed before the first release.
I have also searched for reference to "Magnify", and replaced them
by "Meet".
While updating the backend sources, I have also fixed other parts of
the project, namely:
- Compose file
- Github documentation and CI
- Makefile commands
Introduce CRUD API endpoints for the Rooms and ResourceAccess models.
The code follows the Magnify logic, with the exception that token generation
has been removed and replaced by a TODO item with a mocked value.
Proper integration of LiveKit will be added in future commits.
With the removal of group logic, some complex query sets can be simplified.
Previously, we checked for both direct and indirect access to a room.
Indirect access meant a room was shared with a group, and the user was a
member of that group. I haven’t simplified those query set, as I preferred
isolate changes in dedicated commits.
Additionally, all previous tests are still passing, although tests related
to groups have been removed.
This commit introduces a boilerplate inspired by https://github.com/numerique-gouv/impress.
The code has been cleaned to remove unnecessary Impress logic and dependencies.
Changes made:
- Removed Minio, WebRTC, and create bucket from the stack.
- Removed the Next.js frontend (it will be replaced by Vite).
- Cleaned up impress-specific backend logics.
The whole stack remains functional:
- All tests pass.
- Linter checks pass.
- Agent Connexion sources are already set-up.
Why clear out the code?
To adhere to the KISS principle, we aim to maintain a minimalist codebase. Cloning Impress
allowed us to quickly inherit its code quality tools and deployment configurations for staging,
pre-production, and production environments.
What’s broken?
- The tsclient is not functional anymore.
- Some make commands need to be fixed.
- Helm sources are outdated.
- Naming across the project sources are inconsistent (impress, visio, etc.)
- CI is not configured properly.
This list might be incomplete. Let's grind it.
lebaudantoine