From: Peter Dimov (pdimov_at_[hidden])
Date: 2001-04-11 08:46:29
From: "Doug Gregor" <gregod_at_[hidden]>
> The only hole I can see in my argument so far is that having the "throw"
> could perhaps cause some compilers to produce poor code. Perhaps this is
> enough to warrant using a policy class.
Not a hole. The compiler has no idea what a boost::function points to, so it
must account for the possibility that the call throws.
-- Peter Dimov Multi Media Ltd.