Subject: Re: Windows LISP Interpreter?
From: Erik Naggum <>
Date: 1997/11/04
Newsgroups: comp.lang.lisp
Message-ID: <>

* William Paul Vrotney
| Yet another using the phrase, "toy lisp implementation" without giving
| credit to small Lisp systems.  Seems to me this kind of phasing could
| discourage beginners from implementing their own Lisp interpreters.  Who
| are we to say that people desiring to do this are building a Lisp
| interpreter to use seriously, to learn something or just as a "toy" to
| play with.

I very much doubt that one can _learn_ to program in Lisp well (or at all)
by implementing a Lisp system from scratch with all the incredibly the
hairy stuff that one is supposed to remove from out of one's way in a
high-level language any more than you can learn to drive a car well in a
congested city by building a car by hand-tooling every nut and bolt.

| Furthermore many serious Lisp systems started out as a very simple
| interpreter as described in this thread.  And many of these systems did
| indeed help the Lisp community and fueled new Lisp ideas.  Was the first
| Lisp interpreter a "toy" Lisp?  [etc]

I sincerely doubt that they were built by people who wanted to learn Lisp.
the evidence suggests that they were built by people who knew Lisp well,
even had major designs ready and a reason to build a new system that was
vastly different from starting out on the Lisp journey with ignorance of
the language as their biggest asset.

| Improving an existing Lisp system for a beginner is daunting to say the
| least.

so let's get past the "beginner" stage, then.  frankly, I fail to see the
reason for the admiration you and many others display for beginners and
novices and ignorants.  clearly, it's important to attract new people all
the time, but if we don't find ways to encourage people _beyond_ the
beginner stage, we're worse off than if we don't attract beginners at all,
because we're wasting energy on people who don't stay.

in my view, the catering to beginners and "dummies" in almost all areas of
computing has made the entire field look pathetically simple, and this
means that managers will believe the public impression of programming and
software as trivial tasks, which again means that smart people leave for
(greener) pastures without idiot managers.  in a sense, I think this
fascination with _lack_ of expertise and experience is pathological: it
indicates a fear of either the process of learning or the result of

| Why do think so many Lisp textbooks advocate building a simple Lisp
| interpreter such as Winston/Horn "Lisp in Lisp" as was cited in this
| thread?

note that they do not suggest you go off to write a Lisp interpreter in C.

| Thanks for the suggestion, but although you may not get no immediate
| benefit from it, people building new Lisp systems are needed also.

sure, but people who want to _learn_ Lisp?  I don't think so.

| Someone at Franz should not feel threatened by new Lisp systems and
| ideas.

geez!  what's _wrong_ with you?  why this incredibly cheap shot?

| I would hope Franz's position on this would be to encourage not
| discourage such activity.

_why_?  if Franz, Inc, or any other Lisp vendor, has a position on this at
all, why would it or should it be that they want people to spend four to
ten years learning Lisp by hacking on their own Lisp interpreter when they
could learn it in three months to a year if they got their hands on a real
Lisp system?

even in journalism and at used car dealers, the least respected of all
trades according to some reports, they think experience is good.  somehow,
they don't suggest that people start their own newspapers or dealerships to
"get experience", but encourage people to look at the results of other
successful operations.  why is software so different?

one of the few things psychology theories seem to agree on is that people
need somebody to admire, look up to, some inspiration that others have, in
fact, managed what they would like to do, in some abstract or concrete
sense.  so software people need somebody to admire, too, but what do you
get from visiting even good bookstores these days if not "they think any
random loser can learn to do my job in 21 days"?

my advice is to help people get _past_ the beginner stage and onto real
work.  writing your own Lisp interpreter in anything other than a real Lisp
is to force people to deal with advanced (implementation) issues too early
as well as never getting them out of the _Lisp-beginner_ stage.

if you think this year is "97", _you_ are not "year 2000 compliant".

see for GNU Emacs 20-related material.