>Do you mean remotely as in on a WAN? If so, why are you USEing it? Surely you >should use parameterised queries whether you are using C/S or local tables...

Well, lets examine this. If the tables are VFP - local views are out because you open the underlying table - basically putting you in the same position. Remote views are a pain because of the limitations of the VFP ODBC Driver - stored procs and so forth. Also, my statement is in the context of how most VFP apps tend to be - not necessarily what is ideal. Bottom line, for WAN stuff - SQL/Server will probably be my backend of choice.


>I recall back in the early 3.0 days a number of us did tests with SQL Server >and local tables to compare performance. On 10baseT, as I recall FP came in >faster than SQL Server in these stereotyped tests (about 2X as fast) but >generated 4X the network traffic. That 4X traffic will be the killer on a busy >network or WAN. Since then SQL Server has definitely been optimised as well.

Right, SQL Server is a C/S database - VFP is not.


>But IMHO people are more likely to choose local Vs C/S tables for data >olumes, security, network and reliability reasons, not speed reasons. IMHO >peed is a bad reason to choose C/S without one of those other factors. This >s, of course, not true for Access, Paradox or other F/S systems, which is why >he "party line" is that C/S is faster. With properly optimised tables on a >ecent LAN, FP is often faster.

I agree with you here..

>IMHO performance differences will arise once you have the data. How is it >bound? How is it processed? How many hoops do you have to jump through?

Well... isn't the data the important thing. I mean - if you don't have data, what would you measure??? With regard to binding, I see this as insignifcant as ActiveX data binding can take care of this for you. How it is processed?? That depends on what you are doing and how efficiently you are doing it. How many hoops - isn't application dev. just a series of hoops anyway??


>JR>OK, what sort of HTTP data are you talking about here?
>JP>Application data - stored in VFP, SQL/Server, etc.
>Apart from images and other blobs, why do you want to ship it using HTTP?


Gee.. I have a T1 line and data is sitting on a remote server that a web server has access. I can just get/update data by using RDS - via HTTP. No special WAN here. Pretty cool eh.

Better yet, I have a local Excel sheet that is a template for analyzing data that resides on a web server. How can I get that data? RDS via HTTP. There are loads of reasons I would want to do this.. Too numerous to list here.