>The message that you sent was undeliverable to the following:
> esmoore@flash.net

Where did you get this address? That address is outdated by nearly a year. The UT has reflected my new address for quite some time. esmoore@texas.net.

>Sorry for bothering you via e-mail, but UT seems to be down right now and I have few problems with E-View. I noticed a problem: when I click on Show plan, and then Edit SQL button, the form becomes a mess (seems like SET TALK ON).

EView uses the default datasession, and I was wary of changing universal settings, in case you want to leave EView open while doing other stuff. I could explicitly SET TALK OFF to avoid this, but that might upset even more people.

>However, Show plan gives me:
>Rushmore optimization level for table townsmeth: none
>Using index tag Deleted to rushmore optimize table allfunctions
>Rushmore optimization level for table allfunctions: full
>Joining table townsmeth and table allfunctions (Cartesian product)
>
>I have indexes on MethodID in TownsMeth and on FunctionID in AllFunctions. I also have index on town in TownsMeth. Why I receive a Cartesian product?


Funny, you are the second person to ask about this today. Firstly, VFP often uses intermediate joins to complete a query, the resulting joins of which might require a Cartesian product. It's confusing but irrelevant for us.

Secondly, there are some conditions under which EView will return different Showplan results than you might get running SYS(3054) form the command window. I haven't pegged why it would be inconsistent, and believe me I have tried. For this reason, Showplan might be on the short list for feaature removal in the next version of EView.