Hi kevin,

Also, thanks for this example. It might be a start in comparison.

>>It all looks very reasonable. At the risk of sounding pedantic: you might very well increase VFP's performance for the larger resultsets if you SELECT gender the same way as you do in the C# version- i.e. SELECT just the value- and display the expression in the UI. However, that is possibly taking optimization too far- both versions already seem to yield acceptable performance to create a combined dataset for subsequent reporting.

>Yes, my original intention was never to start a full-blown competition, I was interested to see how C# stacked up, I was never expecting it to be faster, but what I did find was as you said, acceptable performance.

>In fact, the only time C# crumbled was with a first-time query, as you can see it halves the time after that.

>I would love to also hear some suggestions on some data-crunching, perhaps I could do something more after each query, see how much difference there is in data-manipulation, I've heard .Net was awful in comparison to VFP, well lets have a look, I'd be very interested.

The problem with this example is that it does not do much of datamunging. The only real data munging going on here is to fill the age column, which hardly gets significant in the overall picture. After all in your .NET example the data retrieval which takes up the most time is not written in C#, but in VFP.

My challenge posted in Thread #918302 Message #919635, is much more representative to the data munging argument. You get data from the server in various tables in more or less normalized form and the task it to create an overview for displaying or reporting.

The challenge does not provide full VFP code, but rather global directions how you would do it in VFP. If you look at the code carefully, you'll see that there quite a number of VFPs data centric features used, like SET ORDER, INDEX ON, SET NEAR, SET CARRY, INSERT INTO, etc. If these kind of data munging operations can be done in simular performance, with simular programming effort (thus bug risks and maintenance), you'll have gained a believer.

Now, I do not expect you to write a full example, but hopefully you'll see what I mean here with data munging and why VFP is particular strong here. Maybe you can launch some ideas (no, no preprocessing on the databaseserver), constructive comments to get a better picture.

Walter,