Re: RFC 171 (v1) my Dog $spot should call a constructor implicitly

Michael Fowler wrote:
> 
> Regarding the STORE call, if RFC 159 gets adopted, sure, why not.  I'll
> mention that alternative in the RFC's next version.

Cool.
 
> I need to mention the:
> 
>     my int $x;
>     my int $i = 4;
>     my int ($x, $y, $z) = (4, 5, 6);
> 
> syntax in the RFC.  It describes more of the intent behind the change.

Great, yeah this is where it's really powerful.
 
> Sure, the my Dog $spot (@args) looks neat, but it's not any less complex
> than a normal constructor call; if anything, it's moreso. There's no current
> precedent in Perl for such syntax, and I'm loathe to add it without more
> reason.

Yeah, I wrote that on a lot of beer. ;-) The more I think about it, if
we went that route it should look like this:

    my Dog (@args) $spot = "Spot";   # $spot = Dog->CREATE(@args);
                                     # $spot->STORE("Spot");

That is, the args follow the class name. Then it's really just a
left-side indirect object syntax:

   my $q = new CGI (@args);
   my CGI (@args) $q;

   my($x, $y, $z) = new int (@args);
   my int (@args) ($x, $y, $z);

Then I think there is a decent precedent for it. It's an option, at
least. Parsing would be tricky though, at least in the second case.

-Nate