>JVP
>>>Given the choice of technology - one that is language specific and one that is not - all things being equal, I will choose the generic approach.<<
>
>I'm afraid the generalist philosophy has been tried and discarded in almost every sphere of endeavour known to humankind, except possibly medicine. It was Wells who said that "the essence of civilisation is specialisation". He was right. Targeted tools and people are not "better" but they can do a better job because they can afford to focus on their niche and get very good at it. Specialisation is the sign of a maturing industry/profession or individual. The populist "generalist" or "multi-skilling" theory we see today is another "dumbing down" idea to diffuse expertise and keep plodders looking good.
>
>Fact: ADO is SLOOOOOW compared to the dedicated mechanisms in tools like VFP, Delphi and increasingly Java. If you use ADO everywhere, you'd better have a better explanation for the cost you are imposing on clients than the "fact that it is generic" and therefore necessarily better.
>JR

---

John, I agree with you 100%. Especially with regard to multi-skills. I rate VFP my number one skill. I consider myself competent at ASP, Win32, VB and miscellaneous others. JVP's number one skill would appear to be construction of arguments that "look" very compelling. But under close analysis do not stand up. Some of his sweeping statements border on asinine. I personally detest the way he attempts to make opinions look like fact, examples:

"You see, my arguments are not based on technology alone".
"But lets face it, the skill set is not as portable as the combination of VB/SQL".

And so on ad nauseam. Further, he purports to speak from a position of considerable (large corporate) experience - very little of which concurs with my own experience. So little in fact that one would probably need calculus to measure it (as delta Y tends to zero). For example, I regularly deal with large A D O result sets, in fact, the project I am working with at the moment involves taking large chunks of VFP code (predominantly SQL selects) and converting it to ASP using A D O record sets. I am finding in some cases that this is almost impossible - the work-arounds are nightmarish.

My advice to any lurkers is to dismiss the bulk of JVP's comments as witless hyperbole.