All notable user-facing changes to this repository are documented in this file.
The format is based on Keep a Changelog. This repository contains samples rather than a versioned package, so changes remain under Unreleased until a repository release is published.
- Added
zoom_apps/zoom_decisions_compliance_risk_copilot_js, a Zoom App that uses typed decisions to review live transcripts for compliance gaps and risk.
- RTMS media connections now use split mode exclusively. Clients open a
separate data connection for each required media type and send one concrete
media_typeper data handshake: audio (1), video (2), screen share (4), transcript (8), or chat (16). - Generic boilerplates now require an explicit comma-separated selection such
as
RTMS_MEDIA_TYPES=audio,video,transcript. - Purpose-built samples select only the media required for their use case.
- Audio/video streaming samples now manage separate audio and video sockets, including independent keep-alive and reconnection handling.
- JavaScript and Python shared libraries now require an explicit media selection instead of implicitly requesting every media type.
- Unified media socket mode and aggregate data handshakes using
media_type: 32are deprecated and must not be used by new clients.
- Removed
server_urls.allusage from runnable samples. - Removed the
MEDIA_SOCKET_CONNECTION_MODEandMEDIA_TYPES_FLAGenvironment settings from generic sample configuration. - Removed the JavaScript
MEDIA.ALL, PythonMediaType.ALL, and JavaScriptPRESETS.FULL_MEDIApublic helpers.
Replace legacy configuration:
MEDIA_SOCKET_CONNECTION_MODE=unified
MEDIA_TYPES_FLAG=32with explicit media names:
RTMS_MEDIA_TYPES=audio,video,transcriptFor direct WebSocket implementations, use the matching URL and concrete value
for each connection. For example, connect to server_urls.audio with
media_type: 1, then connect separately to server_urls.transcript with
media_type: 8.