[Haskell-cafe] ANN: Elerea 2.x.x + question about AOP in Haskell

[Patai Gergely]

Hello all,

I silently updated Elerea [1] to version 2 about half a month ago, which
deprecated the old interface in favour of the 'experimental' branch of
version 1. The most important change besides this swap is the addition
of the clocked variant of the library, which allows the programmer to
easily freeze parts of the data-flow network, a feature similar to the
clock system of Lucid Synchrone. Also, as of version 2.1.0 (today's
update), the parametric variant was revised and simplified, which is
reflected in the types of the combinators. Feel free to play with the
new features!

On to my question. There are three main variants of the library: a basic
one and two others that extend it in some way. The extensions are
orthogonal -- on the level of denotational semantics they could be
represented as monad transformers (at least I strongly believe so, but I
haven't verified yet). This means that there should be a fourth variant
that combines the two extensions. Also, it might easily be the case that
there will be more independent extensions in the future. This smells
like a perfect opportunity to turn to aspect oriented programming
techniques, but unfortunately I couldn't find a sensible way to exploit
the similarities (of which there are plenty!) and share the core logic
of the library across these slightly distinct variants.

So the question: how does one go about doing aspect oriented programming
in Haskell? Any nice examples in the wild?

Gergely

[1] http://hackage.haskell.org/package/elerea

-- 
http://www.fastmail.fm - Does exactly what it says on the tin