I think firefox tries to print whatever page with the same url was retrieved using a GET request from its cache, and if it doesn't have that page in the cache it does a new GET request. That's why I used the random parameter in my test script to force unique URLs for each request (see my post above).