Re: Different Strokes for Different Folks (Was: Assessing a language)

eifrig@beanworld.cs.jhu.edu (Jonathan Eifrig) wrote
> For each style there is (or should be!) an appropriate
>language. Prolog makes search-based programs easy, but nobody
>would want to write a numerical analysis package in it. Functional
>languages like pure ML are great for manipulating quasi-static data
>structures, but I would hate to write a network-flow algorithm in it.
>
> Surprisingly, there hasn't been much work in developing
>heterogenous programming environments, to support a sort of "mix and
>match" approach to programming. Such tools would go a long way to
>alleviating the language holy wars, I think.


Some new AI* languages (such as POP-11 and Trilogy) take this approach.
Both supply the programmer with the tools of the Functional,
Relational/Logical and Imperative programming paradigms. To the best of
my knowledge an object-oriented mix-and-match does not yet exist.


Part of me finds this approach pleasing, but another finds it awkward. It
would be nice to be able to do prolog-y things in a prolog-y way, ml-y
things in an ml-y way, lisp-y things in a lisp-y way... etc. Where does
one stop? A language which allowed these new language forms to be defined
on top of it would be most aesthetically pleasing, but might not allow for
sufficient optimization on the part of the compiler...


-- Scot


P.S. I don't think trilogy is commercially (or otherwise) availible yet...


==============================================================================


* Don't ask me what an AI language is, or what AI is. AI is probably best
defined as that which cannot currently be done. IMHO, Every AI success
becomes another conventional programming trick. (Dynamic Memory, AI,
Heuristics, etc.) By doing it, we "explain away" the intelligence. I
call these "AI Languages" only since I think it's mostly "AI people" (a
more ambiguous term) who use them or are interested in them.
--

Return to the comp.compilers page.
Search the comp.compilers archives again.