>I doubt that many people code their applications where it checks for the existence of every file, every time it is accessed.

I open every file my apps will access at the top (so my apps never have to discern if it is available). They remain open until a maintenance (pack and index or new map) is requested or the user closes the app. So the idea or practice of willie nelly creating and opening files on the fly is alien to me:o).

Old school flow-charting burn in! I spent nine months in such a school. The first process of every flow chart was to open the tables used in the application and layout the buffer. To understand what files my project would need and how they were to be laid out - the first process to be structured were/was the transaction core of the module or app. When that was complete, I had enough information to layout the tables.

I have noticed not many of do it this way - but I am one of the few who has never had a corrupted tag or FPT - and - I use seek and scan - very few result cursors (if any -except when I need some al SQL data)
I feel out of step - but I am still able to sell a project every once in a while:)