Date: Thu, 15 Oct 1992 14:14 EDT
From: Marco Antoniotti <·······@maya.cs.nyu.edu>
In article <···················@amber.ssd.csd.harris.com> ·····@ssd.csd.harris.com (Brent Benson) writes:
From: ·····@ssd.csd.harris.com (Brent Benson)
Newsgroups: comp.lang.lisp
Date: 15 Oct 92 13:47:47 GMT
References: <·················@CUNYVM.BITNET>
<·····················@news.columbia.edu>
<······················@eng.umd.edu>
Sender: ····@grouper.mkt.csd.harris.com (Network News)
Organization: Harris Computer Systems
····@eng.umd.edu (Charles Lin) writes:
> My question is then: If Lisp is so easy, why is it easy? What
> prevents someone from using, say, ML?
Simple syntax.
I agree, I did a programming project in ML (not a small one: it was a
MIRANDA style type checker for a simple functional language) and
sometimes I got weird type checking errors that were due mainly to
"incorrect" syntax. I ended up parenthesizing pretty much all the code
:-)
Sorry if this is bringing up again a flame. In a recent discussion
appeared in this group, the usual "Lisp is SLOWER than C/C++" flame
raged.
As people often point out, there are a lot of microseconds in a
man-month. If raw performance is not the most critical consideration,
it is far wiser to choose a language that is appropriate to the job.
It is also not true that Lisp is always slower than C/C++. It is true
that C/C++ is often faster, but not always, and often not by a lot.
I just don't buy the claim that pure performance is an issue for most
tasks.
While, as commonsense suggests, languages have their strengths and
weaknesses, I'd like to point out that probably the thing I like most
in Lisp systems is the ability to extend the language and build new
ones via the macro and reader-macros facilities. This is a plus that
no other language I heard of offers (even granting C++, Ada and Prolog
the merit of having redefinable operators).