Oooooohhhh, but that builds up an entire list for the slicing and can damage performance horrendously:

#!c:/perl/bin/perl -w
$|++;

use strict;
use Benchmark 'timethese';

# make it so we split() into many many elements
our $baz = join '', "yes.no.maybe.so." x 5000;

timethese(2000, {
   undef => q{
      ($foo, undef, $bar) = split(/\./, $baz);
   },

   splice => q{
      ($foo, $bar) = (split(/\./, $baz))[0,2];
   }
} );

__END__
Benchmark: timing 2000 iterations of splice, undef...

     splice: 101 wallclock secs (93.86 usr +  0.39 sys = 94.25 CPU) @ 
+21.22/s (n=
2000)
     undef:  0 wallclock secs ( 0.14 usr +  0.00 sys =  0.14 CPU) @ 14
+184.40/s (
n=2000)
            (warning: too few iterations for a reliable count)
[download]