English version
Accueil     À propos     Téléchargement     Ressources     Contactez-nous    

Ce site est rarement mis à jour. Pour les informations les plus récentes, rendez-vous sur le nouveau site OCaml à l'adresse ocaml.org.

Browse thread
(Mostly) Functional Design?
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: 2005-07-18 (16:37)
From: Alwyn Goodloe <agoodloe@g...>
Subject: Re: [Caml-list] (Mostly) Functional Design?
I think some of the problem in this discussion is due to a
gulf between what programmers in industry expect and the
more academic FP community. Lets think back to circa 1988.

OO community was really really small. It had a brand of religiously
committed Smalltalk programmers. Most C programmers found Smalltalk about
as easy to get as they do Haskell. Smalltalk programmers found that the
language and environment made them more productive, but to the average C
programmer it was just too strange to fathom. Then C++ came on the scene.
Virtually no C++ programmer I met in industry during the 1990s actually
used the OO features effectively. Why because all they were given was a
new compiler. The few that did get the new paradigm were taught OO
analysis and design BEFORE they were given a new language.

 About 1991 or so a set of books were published. Grady Booch, Martin
Fowler, and others wrote books about OO analysis and design. These books
didn't speak about code they were mostly about how to do OO analysis.
Examples were taken from application domains such as insurance, banking,
etc. The analysis and design techniques that these authors advocated
eventually evolved into UML/Rational methodology.  Meyer's Object Oriented
Software Construction was a great companion to these less technical books.
It is the analysis and design techniques that make OO useful when paired
with a reasonable OO language like Eiffel or Java. My argument is that the
FP community really doesn't seem to understand that the books and seminars
that Booch and company did in the early 90s (however flawed they were) did
more to move OO from the realm of Smalltalk hackers than maybe even C++.
I hope that there is someone out there that can fill that gap and write
about FP analysis and design.  It seems to me that the FP paradigm is a
better fit with old fashioned Structured Analysis/Design from the 1980s
(Yourdon) than it is with OO analysis/design. Remember dataflow
diagrams....  As someone who likes the paradigm and hopes to see it
spread, I hope some expert in the field addresses this need.

Alwyn E. Goodloe


Ph.D. Candidate
Department of Computer and Information Science
The University of Pennsylvania