>Walter,
>
>For a simple mom and pop app, i'd probably would not user SQL server. Using DBFS possibly with an upsizing strategy would be my first choice.

>Well, you're right on the first sentence, I wouldn't use SQL Server. I'd use SQL Express 2005. <s>

That is entirely your choice, I simply would not. I have to deal with too many mom and pop apps that simply do not justify the trouble of going through any SQL server flavour. FYI, I've got clients in africa running W98 with basically no direct support, no money for support, no internal IT knowledge, etc. Any SQL version would add more headaches to distribution of those kind of apps. There are too many potential problems with correctly installing and running SQL server in combination of other software (that also might run other SQL server versions). If there are problems, simply backup the DATA folder and mail it to me. Right now, it is just one installshield installation and you're up and running. Rock solid.

> >Ease of getting data in and Out ?
>VFP, hands down... USE, COPY TO, APPEND FROM, IMPORT, EXPORT.

>ROFL!!!!!! Too funny!!! Take a look at the XML capabilities in SQL 2005. Look at the options in OpenRowSet. Check out SQL Integration services, and the report builder component of SSRS for exporting data. I think there's WAY more under the hood in SQL 2005 than you may realize.

We are talking about EASE of getting data in and out and not the ADVANCED possibilities and even in that arena I've got doubts (Try to import COBOL data files e.g.). You want to to do a contest of getting a comma delimited file into a cursor or table? How about the following example..

CREATE CURSOR Items (Item C(10), Desc C(200), Price N(8,2))
APPEND FROM (GetFile("csv")) TYPE CSV
SELECT * FROM Items WHERE Item IN (SELECT Item FROM Items GROUP BY Items HAVING COUNT(*) > 1) INTO CURSOR Y
COPY TO MyExcellsheet TYPE XL8

Only 4 lines of code that I can enter interactively in the command window to import, determine duplicates and export again. Even if I need an extra step to connect through remote data through ODBC VFP is more productive.

In many other systems which already use DBFS it is even much simpler. Whether the data comes from DBASE, CLIPPER, FB, FPD/FPW, VFP or another xBase family memember I forgot (And there are HEAPS of those systems), I don't even have to bother importing it, I can work on it straight away. Now try that on straight a SQL server database files you receive from a client. By the time you even have launched the restore option in the enterprise manager, I'll have the results on the desk.

As for your remark on reporting services, what do they offer me what I cannot do already with VFP and Crystal Reports?

>You can use VFPs database engine to solve puzzles in cursors. I've written little short programs to calculate prime numbers, solve sodoky puzzles, number crunching etc. VFP really shines in this area.

>Wow...when can we expect to see marketing literature on that one? <s>

You've missed them. They have been posted on the UT.

>Walter, think back a few hours to what you said about JVP .

Hmmmm, there is not a lot to say anymore. I've not seen him make any intelligent remark about the importance of handling data in a flexible way. He's from a different school who did just not see the light as with many folks up here and as many people who have never dealt with a highly (meta)data driven environment. My conclusions about JVP have been well known (As I told him this directly many, many times personally, and should not come as a shock) and I don't see why you're supprised with this as YOU already knew this. JVP probably was a fine teacher in some respects, but there are others that are taking this many steps to the next level.

BTW, have you ever attended steven blacks presentations on design patterns and meta data ?? If so, do you understand what it is about ??