General talk about software patterns

[Pascal Costanza <pascal AT p-cos.net>]

On 25 Oct 2004, at 17:42, Reich, Shalom wrote:

> On 10/25/04 7:58 AM, "Ralph Johnson" 
> <johnson AT cs.uiuc.edu>
>  wrote:
>
> > It is important to know why great programming languages
> > don't get more market share because until we figure out
> > how to make them win, it is hardly worth inventing better languages.
>
> I know that I only used Lisp in a Programming Languages course.  It
> certainly did NOT give me a feel for the types of problems the language
> could solve and was good at solving.
>
> Speaking for myself, I only thought about the role of programming 
> language
> in framing the solution to a problem when I was forced to change my 
> way of
> thinking from procedural to OO.  The examples in Ruminations on C++ (by
> Andrew Koenig) finally made that explicit in my mind - rather than just
> being a sub-concious piece of baggage.
>
> I still don't know Lisp well but I now see many problems that it can 
> handle
> nicely in the financial modeling world.  There are rumors that my 
> current
> employer uses a private language that LOOKS sort-of OO as a wrapper 
> around
> LISP in order to do Monte Carlo financial simulations.  However, my 
> employer
> (apperently) would rather train people in this private language rather 
> than
> have them write Lisp directly.  I now wonder why this is true.

You don't give enough information here to really tell, but it could 
just be that he uses Lisp in a very natural way. Lisp allows you to 
build the language that works best for the problem domain you are 
working in, and then use that language. No programming language is 
really ideal for each and every problem, therefore providing a 
framework of languages that you can use to tailor your own language is, 
to my mind, clearly the better alternative. This is more or less 
exactly what Lisp provides (among other things).


Pascal

--
Tyler: "How's that working out for you?"
Jack: "Great."
Tyler: "Keep it up, then."