Damian Conway wrote: > or even a arrayed form, when the corresponding index was implicit: > > (@foo,@bar,@zap) := part [/foo/, /bar/, /zap/], @source;
That's kinda nifty. But admittedly, it's not to-die-for necessary, if I'm the only one fond of it. Ken Fox wrote: > and implement classify as a normal sub? Why does everything > have to be built into the first version of Perl 6? Yeah, I agree! Oh, except when it's things _I'm_ asking for. _Those_ are always 100% necessary. :-/ (We're basically asking for everything under the sun, but I think we all know that < 10% of it will actually get in, which is a Good Thing. :-) But sometimes the brainstorming shakes loose something more broadly interesting.) MikeL P.S. As for judging the value of a proposal, I personally try to ask the following questions: 1) Is it a simplification of a universally common but otherwise long/tedious algorithm? 2) Is there only One Way To Do It (Correctly)? 3) Is there a name for the operation so obvious that you can, after being first introduced to it, easily remember what it does? (like "reverse", "split", "while", etc.) Not that I always take my own advice. :-) Other people might have different informal criteria. (For future teaching purposes, I'd love to hear what they are.)
