Rob,
SNIP
>
>I had not previously seen any references to this behaviour in SQL-Select, so there are probably others who would get caught unawares. This situation was somewhat unusual, but should not have been the problem it was. The best lesson to be learned here is to run a test on some sample data before putting into place a critical new technique, particularly if it will be very demanding of the resources of the system, or very time consuming. Bugs and anomalies are not all documented!
>
>This was brought to mind from your comment (above) about faster SQL-Selects when FPTs are defragmeted. It shows that other considerations may also be necessary.
>
>Happy coding...

That was WELL WORTH the length of wind it took you. Thanks for sharing that - I have never heard of it either and I do think it is very significant.

My testing showed some oddity in the processing of FPTs, but this takes it to a whole new level. I'm now thinking seriously of a Wish List item to have more efficient processing and better documentation regarding memo fields and FPT files.

cheers

>
>Rob Steinbach