REMINDER - Contributions to HC&A Report (November 2003 edition)

[Alain Cremieux]

Category : Applications, Groups, and Individuals
Subcategory : Individual Haskellers and their projects

I work in a company making software for payrolling, which is a 
incredibly complex subject in France.
I'd like to use Haskell in my everyday's work, but up to now it's only 
been a personal hobby ("passion" woul be more accurate).
I am convinced that embedded languages ("DSEL") have strong applications 
in software-making companies. Of course Haskell is one of the best 
language for that; but if there are many fascinating papers (resulting 
from hard work & deep insights) on the subject, practical reusable code 
lacks. Most projects are unfinished, and when code exists it is in "as 
is" state, and never reusable without much work.
So as a start I am trying to implement How to write a financial contract 
<http://research.microsoft.com/Users/simonpj/Papers/financial-contracts/pj-eber.ps>, 
a chapter in "The Fun of Programming" by Simon Peyton Jones & Jean-Marc 
Eber. This article details the full analysis (the hard part) of an 
embedded language application. A deriving implementation should provide 
several solutions, depending on design decisions (for instance a 
"shallow" embedding and a "deep" embedding), simplifications rules 
leading to a (moderately) optimised interpreter and a code generator 
(C-- is my target language). And the code should work (calculate the 
value of several common financial options).
Two other key papers for this work are  Compiling Embedded Languages 
<http://www.conal.net/papers/jfp-saig/>, by Conal Elliott, Sigbjorn 
Finne and Oege de Moor, and Programming Graphics Processors Functionally 
<http://www.conal.net/papers/Vertigo/> (Vertigo) by Conal Elliott.
Any advice and/or discussion would be welcome.
Alain Crémieux