"more intelligence in the program" is exactly what you don't need.

The problem you are dealing with relates to the printing of HTML pages, generated by a perl script, using Mozilla/IE. The first thing you should do when trying to address any complex problem, is to split it into the smallest possible "sub problems" such that the action of splitting the problem up doesn't change the nature of the problem.

In your case, the first thing you should ask your self is: "Does the fact that my pages are generated with perl affect the problem?" -- I think you'll find that the answer is no. Mozilla/IE don't care where your HTML page comes from, they'll print it the same if it's served by a CGI, or if it's a file on your HardDrive. So take perl out of the equation: generate a copy of your page that you save to disk, and then tweak it untill you can get Mozilla/IE to print it hte way you want -- IGNORING PERL -- the perl doesn't matter, only the HTML. Once you have the file printing properly, then put the orriginal problem back together: update your perl to generate a page just like the flat html file you have.