Structured concurrency: TaskHandle

[Attila Kelemen]

Correction: I meant " they don't cancel it" instead of " they don't do it".

Attila Kelemen <attila.kelemen85 at gmail.com> ezt írta (időpont: 2023.
máj. 13., Szo, 15:37):
>
> Robert Engels <rengels at ix.netcom.com> ezt írta (időpont: 2023. máj.
> 13., Szo, 14:47):
> >
> > Expecting or using a result after calling shutdown seems invalid. The UI example seems especially incorrect. If I am a user and attempt to cancel an operation and the system says “hey here’s your answer anyway” that does not match the user intent - which is a bad UX.
>
> It depends on the circumstances. Obviously, if showing the result
> would be obstructive, then you don't show it, but there are other
> cases:
>
> 1. You just cleanly show the computation result on a UI is not a harm,
> if the UI would remain visible after the cancellation anyway.
> 2. STS will clearly not manage transactions, and operations often have
> side effects. An API like this can never really promise that a
> cancellation will be honored. In which case it is better not to hide
> the fact from the user, that you actually completed the action, or at
> least know that the process was started, so there might be something
> to inspect. At least I hate UIs that when I cancel the operation, they
> don't do it, then outright lie to me that it was canceled. That breaks
> my expectation even worse, then ignoring the cancellation (which is
> quite often within the allowed set of behaviours).