A Foolish Manifesto

fREWdiculous!

July 2009, DFW.p6m

Today we had another P6M meeting. There were seven of us despite the fact that three of the regulars were gone at a birthday party, so that was fairly heartening.

As you may already know from the Iron Man Feed, s1n did a talk on .WALK, which is a selector based system for introspecting the methods of a class. One really interesting thing about it is that it (apparently?) isn’t actually for dealing with inherited/overridden methods as much as it is for manually tweaking the multiple dispatch that Perl 6 supports.

Just to be clear, multiple dispatch is how Perl 6 chooses what method to run based on the parameters (and invocant) of a method. So you can do something like this:

1
2
3
4
class Frew {
   method foo($self: Int $foo, Str $bar) { ... }
   method foo($self: Str $baz) { ... }
}

And when you call the method it will call the right one based on the params passed to the method. You can even dispatch based on the value of the parameter.

Cool stuff!

  • 0 Comments
  • Filed under: Uncategorized
  • Dallas.p6m: June

    This month’s Dallas.p6m was bigger than before! We had my coworkers Geoff, Neil, and Wes, myself, Graham Barr, Jason Switzer (s1n,) Patrick Michaud, and John Dlugosz. We got a domain hooked up (dallas.p6m.org, which doesn’t point at anything yet,) discussed interesting stories about rakudo optimization (and often lack thereof,) and sometimes delved into perl5 stuff.

    s1n decided to mention that we need to start doing our feature expositions, which is where someone picks a feature in perl 6, does some research, does a talk on it, and then we write some code which is based on it. s1n is going to talk about .WALK, which allows you to look at the Abstract Syntax Tree. I’m pretty excited about that.

    Patrick explained to us how it seems that most traits are becoming declarators. That is, in class foo is bar, class is a declarator, and bar is the trait. Of course the migration isn’t a big deal because defining a declarator is similar to defining a subroutine.

    We also discussed Rakudo’s Unicode support. Patrick mentioned that they are looking for a “golf mode” character which will enable unicode characters for things like >= .

    We also discussed some of the Rakudo release strategy. It’s very similar (for obvious reasons) to Parrot’s release strategy. It’s exciting that they will continue to release regularly (monthly.)

    I can’t recall any other details regarding the meeting. I know we talked about more though. I should start taking notes…

  • 0 Comments
  • Filed under: Uncategorized