I can report with 100% certainty that I've seen SQL Server perform faster than Foxpro to produce reports.

I worked in an AS/400 environment where the report writer on the AS/400 accounting software was a piece of crap. It was set up by our operations department to download the main tables from the AS/400 to our Novell server every night.

I created a reporting system back in the FPW 2.6 days that prompted for some what criteria you wished to use, then built SQL statements with that criteria that produced cursors. The cursors were then used to generate the reports.

One of the last things I worked on at this place was modifying this system, so the data source was SQL Server 6.5 data, rather than local dbfs. All I had to do was change the last part of code that generated the cursor. I had it do a SQLEXEC to Sql Server with the generated SQL statement.

Because I had believed all the talk about the Fox engine being so fast, I first believed I had done something wrong. The longest running reports took about 45 minutes with local Fox data. The SQL Server version took just under 30 minutes. I had to go back and test several SQL statements execution time to verify that SQL Server was indeed substantially faster than Fox on sizeable data sets.

PF

PS. I am now working on a SQL 7.0 system. The system is remotely located and accessed thru DSL. I access the data thru ODBC to VFP and receive the data as fast as local. I have tried and can not do that with VFP data.