Re: exception handling

Well, you can use it that way, but there is no need to do it. If a certain
situation simply deviates from normal, then this situation is in a way
anticipated and normal branching can deal with it efficiently. If you intend
to handle deviations by this mechanism then you hit upon the time
indeterminism problem - and that case has been brought up in this
discussion, as a major cause of geting rid of exception handling entirely.

Nevertheless, I would like to point out that although the execution of
exception handling procedure might vary in time, as long as the procedure
meets a certain deadline it is still OK to use it in hard real-time context.
Cezary


----- Original Message -----
From: "Daniel Simon" <[email protected]>
To: "Cezary Zielinski" <[email protected]>
Cc: <[email protected]>; <[email protected]>
Sent: Tuesday, March 05, 2002 1:59 PM
Subject: Re: [Orocos] exception handling


>
> >Exception handling should be synonymous to handling error situations. We
> >should not try to use that in situations that are not disastrous to the
> >system.                                               ^^^^^^^^^^
>
> I don't agree. Exception handling can be used to handle a situation which
> deviates from a nominal action, thus it can be used to escape from the
current
> action and switch for the new one. Such situations are common, due to the
> environment uncertainty, and it can be an efficient mechanism to handle,
e.g.,
> choosing one way among a set of possibilities to reach a goal in a
changing
> context.
>
> Daniel
>
>
> _______________________________________________
> Orocos mailing list
> [email protected]
> http://mail.mech.kuleuven.ac.be/mailman/listinfo/orocos