A few things I need to point out:
> use strict 'objects': a new pragma for using Java-like objects in Perl
RFC 278 had already supposedly claimed "use strict 'objects'", but this
is flexible.
> =head2 protected
>
> Just take Conway's RFC 188 and do a s/private/protected/g :-)
So you're suggesting that RFC 188 be adopted, but its verbage be changed
if yours is adopted?
> We propose to define a new schema to the C<strict> pragma: C<use
> strict 'objects'>. Use-ing strict objects imports in the calling
> package (or in all packages if declared inside main) three unary
> operators: C<private>, C<protected> and C<static>, that behave as
> stated below.> =head2 private
The importation of functions really causes something to be a module by
definition, pedantically speaking.
Besides, these things seem to me much better suited as attributes, and
I've said that about RFC 188 as well.
Here's what I think: We need a common, extensible attribute system that
allows someone to, via XS, Inline, or something else, twist Perl into
whatever they want. If we made it so that the assignment of an attribute
fundamentally changed the variable it was attached to (which is supposed
to be the purpose anyways, right?), then this RFC, RFC 188, RFC 163, and
possibly others would boil down to a simple case of defining some type
of attribute as having some properties.
We only need two scope declarations: my and our. The rest of this stuff
is attributes that change properties, but not scope. They belong as
attributes.
And yes, I'll RFC the "common attribute system" thingy. :-{
-Nate
