Implementing text-to-speech (TTS) support for Eventyay video streams as part of the multilingual interpretation workflow. I also addressed security-related improvements in the HTTP client used by the SUSI connection layer.
Today, I plan to update PR #15 by fixing the Python cache configuration to match the Eventyay project structure. I will also fix the rough layouting for ui in eventyay videos, bringing transcriptions out of the control panel and have its own box. Reorganize the models to follow the pattern used by other Eventyay plugins, and update the streaming flow so that WebRTC is the primary transport with Server-Sent Events (SSE) used as a fallback. To have an initial connection, we added SSE as primary connection and polling as fallback. would get rid of poling completely now that our signals and middleware are configured. here is a flowchart of current communication: 