Re: Haskell 2 -- Dependent types?

[S. Alexander Jacobson]

I read cayenne.ps and it went somewhat over my head.
I could not get a good sense of when in practice I would really want to use
this.  This is not an objection, just a request for explanation... 

As far as the type system goes, what would be more useful to me is
something preprocessor functionality, like Derive or PolyP, more
integrated into the language.
In one feature, this would:
* eliminate the obnoxiously ad-hoc deriving construct

* facilitate better integration with other languages/systems
For example, it would be nice to be able either to generate a Haskell
datatype from an XML DTD or to generate a XML DTD from a Haskell
datatype. Similarly, it would also be useful to save Haskell datatypes
into SQL databases automatically (I built something like this w/ Derive).

* reduce the need for separate preprocessor systems like HDirect
One could write a type function that would package haskell datatypes for
use by COM and Java components. i.e. one could write code in PolyP that
would convert a Haskell type in to a Java type... without doing through as
much of the preprocessor homework associated with HDirect.  Or the flip
side is that one could construct Haskell types from an IDL again without a
separate system.  (You just need to have a standard FFI as well).

-Alex-

___________________________________________________________________
S. Alexander Jacobson                   Shop.Com
1-212-697-0184 voice                    The Easiest Way To Shop


On Tue, 16 Feb 1999, Lars Lundgren wrote:

> The trend seems to be define wishes for haskell 2, so here are mine:
> 
> We have already accepted undecidable type checking, so why not take a
> big step forward, and gain expressive power of a new magnitude, by
> extending the type system to allow dependent types.
> 
> Cayenne, http://www.cs.chalmers.se/~augustss/cayenne, already has
> dependent types, but seems to be an experimental language only with a very
> small user base. 
> 
> Just a thought.
> /Lars L
>