We implemented basically every version we can get our hands on. Unfortunately it's very difficult to get a comprehensive database of stock files. We implemented the read functionality for exactly this reason, expecting that unknown files would flood in and we can add support as quickly as possible