Excellent example.

if we assume that neither XYZ nor MNO could be updated to be the web app that their customers were demanding and their competitors were providing, then they could have a case for a rewrite, assuming they have analyzed the revenue, cost and profit of doing so. That would indeed be justified.

If either XYZ or MNO were in VFP, however, it would be harder to support such a case because they could move VFP to the web. However if the other non-VFP app had the larger and superior feature set, and could not be web enabled, it might be justified.

Most rewrites which have no justification, but I assumed their would be rare cases where converting from VFP could be justified, but I could not actually think of an example. Your merger example is certainly a potential situation where a conversion might be justified. Thanks for the example.


>Here's one I experienced. Not with a VFP app, but written in a different language. I worked at Company A.
>
>Company A competes with Company B. Company A has product XYZ that has a similar and overlapping feature set to product MNO from Company B. Both companies are marketing the products to the same customers. Company A and Company B merge. Rather than have two products that do essentially the same thing, you decide you need only one. Both are desktop products. Each is written in a different, non-compatible language to the other. Your choices: 1) Take the missing functionality in XYZ and put it into MNO. Discontinue XYZ. 2) Take the missing functionality in MNO and put it into XYZ. Discontinue MNO. 3) Create a new product that has the best functionality of each. Discontinue both XYZ and MNO. In the end, Company A decided to create a new, web-based product. Why? Customers were asking for web-based functionality and other competitors had similar applications that were web-based.
>
>>It certainly is possible for valid businss cases to call for a rewrite, my point was not just that many rewrites are done without valid business cases, but that such a situation is extremely rare when the case is converting a VFP app to .Net. I have never seen a valid justification, and i doubt there could be one, although I certainly allow there could be a case I have not thought of nor seen.
>>
>>>There can be many business cases presented. I once read that on average, an application is rewritten every four versions. There could be valid technology shifts that force that. DOS to Windows, Internet, Mobile, 32 to 64 bit, etc. It could also be due to sloppy/bad design or coding habits. In other words, the code has so much spaghetti, that the company determines its too much of a risk to modify the code. I've seen this happen more than once. Is there a risk? Absolutely. Rewrites always seem to run over budget and over schedule.