David,

What you suggest is probably possible. Howeer, you are falling into the same pit that most FP2.X developers fall into. You are thinking that salvaging the existing code and architectural design is somehow going to save you time in the development of the VFP application.

This could not be further from the truth. Hang around here awhile and follow the threads of folks pulling their hair out as they try to do what you are suggesting. A rewrite is the fastest road to where you want to go.

If you do succeed in doing what you suggest, you still won't have a win32 application. You'll have a DOS application running on a win32 platform, not dissimilar to having a Yugo chasis with a Cadillac body, its still a Yugo.

Take the opportunity to implement all those things you wish you had done differently. The application development process in OO is about 75-80% analysis and design. You already have the analysis done, probably a better analysis than has ever been done before with the length of time the app has been in use. The design will be drastically different as you are moving from character based UI to graphical UI adn from DOS to Win9X, and from structured programming to OOP, and from Monolithic architecture to N-Tier architecture, etc. etc. etc.

If you don't have the OO and N-Tier experience on your team rent it! Hire a consultant to assist you in the design phase of the project and to act as an advisor to your team during the development. The cost of the consultant with be a pitance compared to the time savings you will achieve.

You can choose to ingore what I ahve said if you like. I predict, that if you do ignore my comments and venture into the realm of porting a DOS application to VFP that you will eventually realize that the porting is taking more than twice what a rewrite would take.