Hi Jan,
For many years now, I've been thinking that it would be good to
rearrange the way that Hugs works so that the user interface to
the interpreter is written/scripted entirely in Haskell using an
library of primitives to call the underlying Hugs engine. The
distribution would include a standard interface that most people
would use unchanged, but those with special needs like yours
would be able to modify and customize it as necessary. In some
cases the read-eval-print loop might disappear altogether. I
believe that something like this will be possible when the HEP
(Haskell Execution Platform) is available.
| As an aside, a way to make permanent variable bindings would be
| great too. However, this conflicts with the script vs. session
| philosophy as laid out in the 'The implementation of the Gofer
| functional programming system' by Mark P. Jones. HBI has it, so
| why not Hugs?
This one also comes up quite frequently. It's difficult to do
only because it's difficult to know what should happen to local
variable bindings when you start switching between modules,
reloading, etc. As far as I know, nobody has come up with a
good way to address this problem. To be honest, I've never
actually missed this feature when I use Hugs. The editor is
never more than a couple of keystrokes away, and I get a
persistent record of all the definitions that I've made, which
I can edit and reload as often as I like without every having
to wonder which versions of my definitions are in scope. But
I know other people think differently. I guess it's a culture
thing: it depends what you're used to ...
All the best,
Mark
