Re: JACAL, scm

[Evan Prodromou]

>>>>> "RB" == Rob Browning <address@hidden> writes:

    RB> The problem, I think, is that Tom wants the #f '() equivalency
    RB> so that guile can handle common-lispy/elisp/SCM code without
    RB> much modification.  He also wants (and I presume others do
    RB> too) guile to be able to execute elisp and guile code in the
    RB> same process with the functions closely exchanging data with
    RB> each other.

    RB> It's the exchanging data part (if you want it to be as
    RB> transparent as he does) that seems like it might require a
    RB> more fundamental change to guile.  For the other languages
    RB> like tcl, I don't think people have ever really been talking
    RB> about such tight, transparent interactions.

If I'm not wildly mistaken, it seems to me that there's not really
much of a design or even a strategy _at_all_ for the somewhat lofty
idea of having multiple languages compile down to Scheme, and then run
from guile.

Given that, #f/'() seems like a fairly small item indeed. If I was
going to say anything on the subject, I'd say that the more
standards-compliant we are with the Scheme implementation, the easier
it's going to be to have a consistent, reasonable, and practicable
translate-language-X-to-Scheme design.

~ESP

-- 
Evan Prodromou
address@hidden