my $line = 0;
my $file = "file.txt";
open (FILE, $file) || die "Could not open $file\n";
while (<FILE>)
{
      print "I am on line ", ++$line, "\n";
}
close (FILE);
[download]

"More specifically"? What's wrong with using $.?

There's nothing "wrong" with using $. since it accurately represents the line number read of the last file-handle read. You can even track more than one file at the same time:

open (A, "ls|");
open (B, "ls|");
while (<A>)
{
        print "A: $. ";
        <B>;<B>;
        print "B: $. ";
}
close (A); 
close (B);
[download]

Which returns the output:

A: 1 B: 2 A: 2 B: 4 A: 3 B: 6 A: 4 B: 8 A: 5 B: 10

However, $. is what I would call a "magical" Perl variable because it comes from nowhere and has no inherent meaning. With English, it is $INPUT_LINE_NUMBER, and with IO::Handle you can use the method input_line_number, but the $INPUT_LINE_NUMBER variable is still "magical", and the IO::Handle system just makes things more difficult to implement.

"Magical" variables, or those that are effectively, though not intuitively, linked to another variable or action are certainly able to be used, but the issue I have with them is that to the average programmer, especially one not intimately familar with Perl, is that they don't make any sense. Using them on a quick hack is one thing, but using them on a program that might have to be maintained by others is a counterproductive form of obfuscation.