Rick

>>This means for somethings that we can do with SQL or data commands in VFP you'd have to manually parse the XML data (or DataSet) by running through the data in a loop and manipulating. Still this is simple and doesn't add a lot of code or speed overhead especially if this sort of thing gets abstracted into classes where it belongs.<<

As you suggested it most certainly does make a difference if you have large lookup tables. We currently deliver a 13Mb VFP executable with some monster tables in it. We can process 1000 of a certain document type per minute with complex parsing and lookups using VFP. I'm absolutely amazed how quickly LIKE() statements run against huge local tables. We have abandoned the dotNET version completely because it was taking around a second per document. Some work was done to try this using indexed SQL Server tables rather than XML but the chaps were dismayed at the work required and we won't be trying again till the next release of dotNET.

Regards

JR