>And the beat goes on ...
>
>I'm one of those who think dbfs still have their place and that VFP remains viable in some circumstances in 2007. I suppose I may be in denial despite having used SQL Server/Oracle exclusively for customer data since 1995... ;-) but in that case, why are we seeing all these free "lite" databases from major vendors who sell proper databases? Perhaps they're in denial too. ;-)
>
>Here's why I think dbfs will survive: they're everywhere; they're cheap; they don't require expert maintenance; they don't need big hardware; and a sizable market seems immune to every argument against their use. Recent news that 70% of businesses are not planning to move to Vista for reasons that include reluctance to upgrade hardware and compatibility with existing software makes me wonder whether that sizable market is even bigger than I thought.
>
>If I were setting out today to sell an IT product, I'd target the small business market with apps that will work on the lousiest hardware, give me and the customer maximum control, require as few external licenses as possible, and require as little maintenance as possible. If I could achieve that with a lite C/S product, I'd do it, just to get the tick in the box. If the customers are likely to use Quickbooks (which is very likely) I'd lean towards SQL Anywhere since that's what Quickbooks uses. Otherwise I'd use dbfs or Jet with VB or VFP as a front-end.
>
>What would you do?

I think in an earlier post you made a very important point in recognizing that one of the impediments to discussions of 'best practice' or the strenghths and weaknesses of Fox, .net etc is that in our profession ( or art form ) we attempt to solve an almost infinite array of business problems for a wildly diverse client base with very eclectic skill sets.

So of course any 'solution' must take into account horses for courses.

That said, first there is the question of what bag of tricks will give you the most flexibility in solving the problems your own business model is likely to present to you and what is best ROI in learning new stuff.

I do all my VFP work using VFE / SQL Server. The VFE framework has had me writing n-tier apps from the beginning that are entirely view based. Dynamic view parameters ( the SQL remote views sql statements are built on the fly with params that are not used being ignored ) makes it very flexible for working against a SQL back end.

It is no harder to use SQL than to use DBFs in this model, so I choose to use SQL, since that gives me the advantage of a real CS database. Jim Duffy and Gary DeWitt often made the point that using MSDE/SQL let you have one code base for a simple desktop app or a much larger installation. Especially useful in verticals.

Clients seem a lot less nervous about VFP if they know their data is in SQL. I also find it nice that with my major clients since their data is in SQL if they come to me wanting something VFP does not do comfortably I can point .NET at the same data a supplement the VFP app. Seems to me the best of both worlds.

I know clever people have pushed VFP beyond what it should reasonably be able to do but I would bet that if Rick Strahl was starting a big web project the first tools he would reach for would not be VFP, DBFs and Westwind. ( and he would be quite capable of doing the job with those tools )

I'm starting to use Strataframe as my entre into .NET. I find it philosophically similar to VFE and I don't see data handling as a major problem. Business objects, objectified data and a framework that handles some of the problems a beginner doesn't even know are there is a nice way to get started in the right direction and seems pretty cost-effective to me.

I think currently writing VFP apps that are DBF table based ( no views ) binding directly to data and using a file server model is limiting and the skills involved are not very useful in any other development environment.

It also seems that in the .net world a lot of beginner books etc stress how to do stuff the 'getting started' way and learning stuff that won't scale.

Learning to ski by snowplowing is comforting - but it isn't pretty - and it isn't skiing.