I use something like that every day -- I have a script called notch.pl that outputs a blank line and the time after a new line of output arrives if it's a different minute that the previous line's time.

Here's the code:

#!/usr/bin/perl -w

#  Notch out time when lines from 'tail -f' arrive so that lines
#  from different minutes are separated by a timestamp.

use strict;

{
    my $lastMinute = (localtime)[1];

    while(<>) {

        my $thisLine = $_;
        my $thisMinute = (localtime)[1];
        if ( $thisMinute != $lastMinute ) {

          $lastMinute = $thisMinute;
          print "\n";
          print scalar localtime;
          print "\n";
        }
        print $thisLine;
    }
}
[download]

I probably could use File::Tail as merlyn suggests but just threw it together because trying to readh read the tail of a log file is head hard enough -- this just breaks it up a little.

Based on the amount of traffic that I see, I don't worry too much about it using an extra pipe or process.

Update Correct two typos .. the first of which I made again, and had to go back and fix again.