When is a requestScope variable actually destroyed?

[dkarr]

I know what "request scope" means in the context of web applications, and I'm sure that Web Flow implements it the same way as other web frameworks. However, I'm trying to rationalize this understanding with the statement that I see in the documentation, which is:

"Request scope gets allocated when a flow is called and destroyed when the flow returns."

When the flow returns? Does the end of a single request correspond to when the flow "returns"? What exactly does this mean?

[InverseFalcon]

So you press a button on a page. This might start a new flow, or resume a flow already in progress. Either way, request scope gets allocated. Server-side processing occurs, and eventually a page needs to be rendered. When the page gets rendered (or if a redirect occurs), then request scope is destroyed. So it's still for the duration of an HTTP request.