Currently a user has to jump between the ComfyUI and ComfyStream interfaces to develop a workflow, and validate that it works as a real time video workflow. This is a cumbersome experience, and slows down development time.
A more ideal user experience can be as follow:
- A user launches a SINGLE application, which can toggle between the current ComfyStream front end app and the ComfyUI front end.
- A user would develop the workflow in ComfyUI. Once ready, hit a "Load Real Time Video" button, which automatically exports the workflow, loads it in ComfyStream, and starts the stream. This allows the user to validate the "live video mode" of the workflow.
- This flow eliminates the need for the initial screen of ComfyStream, removing a whole step in the user journey.