Milestones
List view
We should be able to have blocks like: - Note point - Scripture reference - Slide content (which may also be a scripture ref) Then there doesn't need to be a separate slides viewer, and the TTS can read out any text. Should slide content get mixed into the same stream, or should it go in a special pinned area or something?
No due date•3/3 issues closedWe're fighting the complexity monster. Mostly successfully, but there's some ugly stuff in there too.
No due date•5/5 issues closedRather than having Proclaim output a video feed and then capturing it and transmitting it over a side channel, we could directly watch the slideshow status via the API, pull the corresponding text from the show itself, and show text. Resources: [validate-proclaim](https://github.com/kcarnold/util/blob/master/src/validate_proclaim.py), and I have a watch script WIP in my sandbox repo. -- I would like to pull in the current slide text from Proclaim. I've written the beginnings of a script to handle this (proclaim_client.py), which polls Proclaim for the current slide and looks up the text. I'm thinking of actually pulling the entire current presentation into Yjs (as an array of items, each of which has an array of slides). The proclaim client will need to watch for presentation updates (since that can occasionally happen live) and on-air-status (current slide) updates and update the Yjs accordingly. The client will render that in a way that I haven't entirely designed (it'll replace the "video" element eventually but not yet), for now I just want a proof-of-concept that we can get and update the current slide text on the clients. Architectural question... this will need to run on the computer that runs Proclaim. Lowest-hanging solution is I check out the repo there and use uv to manage a Python environment. But how could I have it keep itself up to date? I also want that same app to be able to run the speech transcription (so I can script it to start instead of having to manually start it on the web site and approve permissions every time. Will we be able to do that with the Python script, or will I need to make this a native app of some sort, maybe Rust/Tauri or Swift, or Electron?
No due dateTranslated text gets read out loud. Currently under development on the `tts` branch. So far it works pretty well, but we should have it read automatically (if "playing" / auto-advance). Perhaps there's a playhead cursor that tracks what's just been read.
No due date•1/1 issues closed