At 1:54 AM +0200 6/19/05, Juerd wrote:
> Except for attributes, which play a different game: the colon comes
> *instead* of the dot as the twigil, while the accessor method gets : in
> front of its name. If I recall correctly, the syntax is very misleading
> in that it is NOT part of the name.

I would argue that the same rule applies to attributes and methods 
alike; privates of both have ':' as part of their name, and *every* 
attribute or method reference, public or private, has a '.'.  No 
"instead" here.

So if you just factor out the ':' as being part of a private thing's 
name, then all other discussions for syntax can treat public and 
private the same way.

This goes for both explicit invocant forms like $obj.attr or 
$obj.meth and implicit forms like $.attr or $.meth, where in all 
cases 'attr' or 'meth' may or may not contain a leading ':', but it 
doesn't matter for the main discussion, except that ':' can't be used 
in the trailing edge of the operator.

Its much more consistent that way.

>      $.bar      $.:bar      # I wouldn't mind if $:bar were an *abbreviation*

I agree and would accept $:bar *if* it were officially an 
abbreviation, as you suggest, just like one can omit the '.' on 
things like $foo.[0].{'bar'}; however, the full/longer form should 
always be valid syntax.

-- Darren Duncan