This example uses background tasks to concurrently generate and stream lorem text, simulating how an app might make multiple API calls and display the results as they become available.

The state keeps track of several dicts that are keyed on the task id:

• running: bool if the task should keep processing data
• progress: int how many iterations the task has completed
• end_at: int the task stops after this many iterations
• text: str the actual generated text

This is the background task that does most of the work. When starting, if a task_id is not provided, it assigns the next available task id to itself; otherwise it will assume the values of the given task_id.

The task then proceeds to iterate a random number of times, generating 3 lorem words on each iteration.

The page initially only shows the "New Task" button. Each time it is clicked, a new stream_text task is started.

The tasks are presented as a grid of cards, each of which shows the progress of the task, a play/pause/restart button, and a kill/delete button. Below the controls, the text streams as it is available.