Ilya Zakharevich wrote:
>
> $_ is not ALLCAPS. @EXPORT_OK should die (see RFC 233). @ISA is on
> its way to its grave already, see C<use base>.
Yeah, but you're still just sidestepping my point. Your position seems
poised on the hope that no more special variables get introduced, or
that some of the existing ones go away. To be effective, I think your
RFC needs to create a fundamental different approach to all these
things. Otherwise you're just creating inconsistencies.
As for the different namespace for subs thing, I don't think that's the
way to go. You do have to call special subs directly, at the very least
from other special subs. And what if you wanted to provide an OO and
tied interface?
sub STORE {
# store data
}
*store = \&STORE;
If I call $object->store, what will happen if STORE is a special method
like you propose?
I personally think the best solution to this issue is not a technical
one, but a conventional one. Let's pick a convention - like one of the
ones I mentioned in my previous email - and stick to it, plain and
simple.
-Nate
