I recently started on a project that involves migrating some data from a legacy app & database into Drupal. The old application is a collection of PHP scripts that basically just generate forms, accept data, insert said data into the database, and output it on a website. Pretty simple stuff - there's not a whole lot going on. It was developed long before many of the popular CMS's and frameworks came to be, and probably before people really started paying attention to the character encoding of their data.