Jay,

I think so. My understanding is that most of us use standard optimization features and only tinker further if performance is inadequate and the customer can't "fix" it by buying a better box.

In this case, there was very little to optimize. SET TALK etc are standard. APPEND FROM rather than SCATTER/INSERT is also fairly standard, but in this case it made a whole millisecond difference. I posted some other code where using APPEND is an order of magnitude quicker than SCATTER/INSERT. Just not in this example.

I guess I was responding to the suggestion from elsewhere that the above "tweaking" is inappropriate. Nonsense- those suggestions are perfectly normal in a VFP app. However, if we had to start coverage profiler and experiment/tweak in detail to match performance, that would be a trick shot as well.

As it is, we are advised that after a reboot and a good night's sleep, the differences were actually negligible. So those others who pounced gleefully on the example and began announcing sensational conclusions, ought to be ashamed of themselves.

Regards

j.R