[plt-scheme] The perfect teaching language--Is this too much to ask for?

On Sun, Jun 14, 2009 at 5:21 PM, Noel Welsh<noelwelsh at gmail.com> wrote:
>
> The issue is one of usability. It is pretty easy to get ML to spit out
> an incomprehensible type error  at a location far removed from the
> original error, and ML's type system is fairly simple.

I wonder, is that an insuperable problem, or simply one no one has
cared to solve because ML is not a teaching language and the feeling
is that users can learn?

> Furthermore, I
> believe the experience of the PLT crew and others is that allowing
> students to experiment with their programs -- actually run them
> despite type errors -- leads to better understanding.

That is a good point; I just feel that ML comes closest to mirroring
what is taught in datastructures and algorithms courses, without much
impedance mismatch. It's far more "executable pseudocode" than pascal
was, in that way (or maybe pseudocode has just become more
ML-influenced over time :)).

martin