[boost] Re: Re: Re: Re:Re:Re:Re:AttemptingresolutionofThreads&ExceptionsIssue

From: Eric Woodruff (Eric.Woodruff_at_[hidden])
Date: 2002-08-13 14:18:52

I'd hate to start a different argument, but exceptions aren't always used to
indicate an unrecoverable situation, or even a situation that needs
recovering at all.

Though, not a very good example (because it depends on subjective design
tastes), consider method in a graphics program that asserts that a line and
triangle do not intersect because this is a necessary condition. The best
piece of information to throw might be, in fact, the point of intersection.

----- Original Message -----
From: William E. Kempf
Newsgroups: gmane.comp.lib.boost.devel
Sent: Tuesday, 2002:August:13 3:03 PM
Subject: Re: Re: Re:
Re:Re:Re:Re:AttemptingresolutionofThreads&ExceptionsIssue

----- Original Message -----
From: "Eric Woodruff" <Eric.Woodruff_at_[hidden]>
> So, it seems exception hierarchies are not feasible unless the abstract
type
> provides a clone () method. We could possibly implement cloning wrappers
for
> std::exceptions and then unwrap them at the other end.
>
> For user exception hierarchies, they would have to follow the same clone
()
> semantics.
>
> How often does a user need to support entire exception hierarchies across
> threads?

Is it ever useful to know what the exception type is in this case?  I ask,
because it seems to me that there's no way to recover in this case, and the
only reason for propogation is to give some possibility for a clean exit
instead of an abrubt halt.  If you can't recover, then what need is there
for knowing the exception type?

Bill Kempf
_______________________________________________
Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost