Stability: VFP8 versus VFP9 (by John Koziol)

I’ve gotten some emails from old friends and VFP developers where the sender is asking me which version is best to upgrade a legacy application. Some folks are worried that the “newness” of VFP9 may make it a less stable upgrade whereas VFP8 has had a service pack and been around a while.

So far, it appears that customer reported bugs for VFP9 are 70% less in the same post-RTM timeframe than VFP8 bugs. And we have had roughly the same amount of beta testers and early adopters between both versions.

The trick is that we had many more tools to ferret out weaknesses in the core code of VFP for the VFP 9 release than ever before. Even with the huge changes to reports and the SQL engine, we had test coverage that could find almost any hole in the application.

To me, clearly, VFP9 is the way to go. I have no reservations in saying that this version of VFP is by far the most stable ever released.


Comments (7)

  1. I agree with you John. Having been in the Beta and having used VFP9.0 in a production environment I concur with your comments.

    <a href="">Alex Feldstein</a>

  2. any chance that we will ever see full Unicode support in fp??

    My favorite genealogy program The Master Genealogist, uses vfp for the platform which was great until I got documents in Russian an d ukrainian that I can’t record properly.

    have requested on UT for both VFP 8 and VFP 9 cycles. will request again for VFP 10.

    Thanks Douglas

  3. John Koziol says:

    What do you mean by "record"? Are you sure this isn’t an application problem and not a VFP problem?

  4. What I mean is that there is no good way to store Unicode data in VFP currently. The dataset must be defined by a code page when designed.

    While 1251 works fine for most of the european languages it doesn’t natively support cyrillic.

    So what I mean by record is there is no effective way to support multiple code pages within the VFP enviroment.

    Better yet would be to natively support Unicode as a dataset code page. or at the very least allow for code page selection at the record level.

    thanks douglas