Level Extreme platform
Subscription
Corporate profile
Products & Services
Support
Legal
Français
Performance differences between VFP 5 and VFP 6
Message
 
To
All
General information
Forum:
Visual FoxPro
Category:
Troubleshooting
Title:
Performance differences between VFP 5 and VFP 6
Miscellaneous
Thread ID:
00123128
Message ID:
00123128
Views:
59
We've got a very large application that supports multiple instances of the same forms. A few of these forms have a large pageframe with about 18 pages, with anywhere from 10 to 50 controls per page. These controls are all contained within a container class that is not instantiated until the first time that the user goes to the page. The EXE was built using VFP 5.0.

If we open 5-6 of these forms, and then go to every page on every form, the time that it takes to change pages becomes essentially unusable, anywhere from 45 seconds to as much as a minute or two. As you open more and more of these forms, the performance continues to degrade further and further.

Note that if you open 5-6 forms, but leave all but 1 of them on the first page and only cycle through the pages on one of the forms, performance is still not as good as when just one form is open, but it is acceptable. So this performance degradation appears to be directly related to the number of objects that have been instantiated. This has been tested on machines with 64+ MB of RAM, so this is not an issue of getting into virtual memory.

Under VFP 5, this same executable runs with completely acceptable performance, even with twice the number of forms. Changing pages takes at most anywhere from 2-10 seconds. It would appear that some sort of change was made between 5.0 and 6.0 that causes 6.0's interface performance to degrade much more rapidly as more and more objects are instantiated. If you run this same EXE side by side under VFP 5 and VFP 6, with only one form open you may not notice any difference. As you open more forms, and cause more and more objects to be instantiated, the difference becomes very obvious very quickly.

From preliminary testing, building the EXE under 6.0 doesn't seem to make any difference in performance. An EXE built with 6.0 exhibits the same problem as one built with 5.0. I'm still working on coming up with a small sample that will demonstrate the problem. If I can do that, I'll post it here.
Next
Reply
Map
View

Click here to load this message in the networking platform