Re: [xml-dev] Formalism and complexity

Michael Champion wrote:

>  It's hard to argue with the proposition that logic, set theory, and a 
> deep understanding of RDBMS dogma (or OO dogma, for that matter) 
> *should* help one build successful complex software systems.  
> Unfortunately, I don't know of much evidence that it really *does* help.

It would probably work very well if problems remained constant throughout a 
project (like, say, building a bridge across a river), but in high tech, 
they do not -- we have only a limited need for people who can create an 
algorithm to do a computation in Olog(n) instead of On(log(n)), but we have 
an enormous need for people who can track changing requirements and 
userscapes and refactor code violently and continuously to match them, and 
we have an even bigger need for people who help build consensus and 
communities of users.  It's mostly flexibility and unscientific grunt work 
that brings success.


All the best,


David