K, Dylan, etc etc etc.

Raul Deluth Miller rockwell@nova.umuc.edu
Fri, 10 Feb 1995 13:44:41 -0500


Rainer Blome:
   yes, jvs, we should design our own language.  but of course we
   should check whether existing languages fit our purposes.

   k doesn't, it seems write-only and unsuited for systems programming.

The problem with K isn't that it's write-only (it's not).  The problem
with K is that, for "our purposes", it might as well be vaporware.  [I
don't know how long till it's publically released, but I suspect that
will be something like a year or so.]

Also, I was presenting K as a model for the HLL (the applications
programming language), not the LLL (the systems programming
language).  Please be aware of the distinction?  [On the other hand,
since K allows for run-time linking with low-level code, the
distinction between HLL and LLL is somewhat blurred -- but in the
sense of being able to do low-level things in the HLL and not in the
sense of the HLL being unsuitable for high level things.]

-- 
Raul D. Miller
Account expiring on 13 Feb'95: <rockwell@nova.umuc.edu>/<rockwell@nova.umd.edu>
Temporary alternative:           <rdr@home.merit.edu>/<rdr@legislate.com>
Temporary alternative:     <raul@tad.micro.umn.edu>/<raul@gopher.legislate.com>