I agree with you on all fronts and thanks for the post

Our application is data analysis / BI oriented and lives and dies on Rushmore.

Sybase's Advantage is top on our list of databases due to it's Rushmore compatible indexes and large table size and has a .Net provider.
We looked at the VistaDB engine, great and all that, but nowhere near fast enough.
We still need to benchmark the Ado.Net to SQL Server 2008 round trip versus the VFP Rushmore lookup.

It is clear that our application architecture under VFP will not fly under ADO.Net due to the .Net all-data-in-memory dataset paradigm versus VFP's data on disk paged into memory paradigm.
Then we still have the issue of passing large amounts of data via COM Interop - another performance bottleneck.

So it is very clear tht We will need to change our architecture to passing around as small datasets as possible.

Another good reason for the Move to .Net is that overtime VFP will become less and less compatible with newer versions of SQL Server and more importantly how much longer before VFP's UI "breaks" under the GUI of future
windows versions, be it Windows 8, 9 or 10.

Had MS not abandoned VFP many years ago, there would be no need for this thread.

We cant ride the VFP scooter forever, you got to get off some time, better slowly and at a control pace than to be thrown off by MS. - Just my opinion



>Well,

>I was not the only one with doubts. Speaking mostly from the technical side of things,
the vfp approach of having large amount of data locally available makes some tasks easier.

>The C/S approach has much going for it from safety and areas as well and the DotNet
mantra is the disconnected data set held in memory. Advantage allows you to access vfp tables
from Dotnet and might give a more vfp-like way of handling data in your Dotnet app.

>But weigh the needs of keeping the same approach vs. a more C/S spirited one
not only on the structure of the current app. On some points the newer "biz models"
are miles ahead of the simple 1 tier vfp app (and IMHO better than simle MVC designs) -
it is just that MS missed a couple of other things in appreciating the use of local data.

>At least recheck and perhaps rearrange the architecture of the vfp app -
otherwise IMHO a rewrite is only cosmetic and probably a waste of resources.

>knowing that I am guessing a lot on almost no facts (see TC's post)