True. And (for the first time here:), I actually knew that.

However, as I understand it, when I defined my @list, I actually defined a typeglob named *list which contains one each of a scaler, an array, a list and a hash all named list (and accessable locally with the appropriate funny character prefix).

Moreover, when one iterates over @list using one of the foreach or for...in constructs, the scaler $list is automagically (sorry. I forgot the correct term) set to the next value from @list at each iteration. Hence, my attempt to use that.

The problem in this case (thanks to jeffa!), is that the iteration in being performed outside of my scope, so $list is never being set.

At least that is my understanding at this point. Maybe someone will correct me if I have this wrong.

You have that wrong. :) Witness:

my $list = 'some default item';
my @list = (1, 2, 3, 4);
foreach (@list) {
    print "$list\n";
}
[download]

If you change the print line to print "$_\n";, you'll see the aliasing in effect. (It doesn't matter whether you use for or foreach, as they're synonyms.)

Of course, there's a deeper misunderstanding, namely that declaring a lexical does not create what most people in the know might consider a typeglob (at least, not in a symbol table. It's put in a pad.). Nor are the values pointed to by typeglob slots automatically populated by iterating over an aggregate of the same name.

It is true that when you do cause a new typeglob to be created, it does automagically get a new scalar in the appropriate slot, but that's done as an optimization, and not for any aliasing.

Proof if ever it were needed that a little knowledge is a dangerous thing

Just as I think I can run, I get sacked!

Thanks for putting me straight and my apologies for contradicting those that know better than me.