Timeout and Waiting Strategy

Visualization boundary: timeout diagrams explain waiting boundaries, not static flow structure.

TriggerFlow is an event-driven runtime. A call to start() is not guaranteed to produce a result immediately.

1. Timeout timeline

How to read this diagram

• get_result(timeout=...) and get_runtime_stream(timeout=...) are separate waiting paths.
• A stream timeout only stops side-channel consumption. It does not terminate execution.

2. Common waiting sources

• when() waiting for external events
• pause_for() waiting for continue_with()
• runtime stream waiting for new events or stop_stream()

3. get_result(timeout=...)

When execution.get_result(timeout=...) or flow.start(timeout=...) times out, it returns None and emits a warning.

That usually means:

• there is no final result yet
• you may have forgotten .end() or set_result()

4. Stream timeout

get_runtime_stream(timeout=...) timing out only stops stream consumption. It does not end the execution.

5. API boundary guidance

• always define explicit timeouts at web/API boundaries
• use start_execution(..., wait_for_result=False) for long tasks
• do not encode “wait forever” inside business chunks

6. Recommended usage

• linear flows: direct start()
• long-lived flows: start_execution() plus external waiting
• UI intermediate states: runtime_stream