Re: globs?

--- Larry Wall <[email protected]> wrote:
> On Tue, Apr 08, 2003 at 06:18:09PM -0500, Austin Hastings wrote:
> : Larry, for the fourth time it has occured to me that saying:
> : 
> : class MyClass {
> :   has $.attr;
> : }
> : 
> : implies being able to say:
> : 
> : class MyClass {
> :   has $attr;   # no "dot"
> : }
> 
> I don't think it has to imply that.
> 
> : What do you do when that happens? (Is that the way to do class
> vars?)
> 
> No, if it means anything, it means an attribute that doesn't have an
> accessor method.  But at the moment I don't think it means anything.
> 
> Maybe it's irrational, but I got sick and tired of looking at methods
> in C++ and not knowing whether "foo" was an attribute or a variable.
> So I'm not eager to put the same ambiguity into Perl.  Secondary
> sigils in Perl 6 are intentionally aimed at making scope and lifetime
> distinctions visually explicit.  More specifically, I'm trying to
> make a distinction that many natural languages make between marked
> and
> unmarked constructs.  It's the unusual scopes and lifetimes that
> should
> be considered marked.  Ordinary variables don't need to be marked.
> It's just a funny kind of Huffman coding, from a linguistic
> viewpoint.

That's a good point, and a passable explanation. The distinction
belongs in the docs, so I'm rerouting this message to p6dox.

MikeL, since A12 isn't here, there's no object/class stuff going, is
there? 

=Austin