Newsgroups: comp.lang.scheme
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!udel!gatech!howland.reston.ans.net!news.sprintlink.net!siemens!princeton!news.princeton.edu!blume
From: blume@dynamic.cs.princeton.edu (Matthias Blume)
Subject: usefulness of higher-order features [Re: GEL #f/nil, SCM choice]
In-Reply-To: tammet@cs.chalmers.se's message of 13 Nov 1994 16:11:14 GMT
Message-ID: <BLUME.94Nov13125852@dynamic.cs.princeton.edu>
Originator: news@hedgehog.Princeton.EDU
Sender: news@Princeton.EDU (USENET News System)
Nntp-Posting-Host: dynamic.cs.princeton.edu
Organization: Princeton University
References: <Cz0ADt.Gv@viewlogic.com> <39r0j0$han@nyheter.chalmers.se>
	<OZ.94Nov10230218@nexus.yorku.ca>
	<BLUME.94Nov11103530@dynamic.cs.princeton.edu>
	<3a5dr2$amb@nyheter.chalmers.se>
Date: Sun, 13 Nov 1994 17:58:51 GMT
Lines: 37

In article <3a5dr2$amb@nyheter.chalmers.se> tammet@cs.chalmers.se (Tanel Tammet) writes:

   The term "C-like scheme" is fairly vague, of course. Eg, if you just
   limit higher-order stuff, is your code then properly "C-ish", so to
   say? Suppose it is. Then I do personally see a number of strong
   reasons to prefer even this C-ish scheme to C. I have hacked lots
   of code in C and lots of code in various lisps and schemes. I am
   _not_ willing to write non-toy applications in C. For me program
   development in C is several times slower when compared to scheme,
   though I have personally no inclinations to use higher-order programming
   style except simple cases like some, every, member-if and the like. 

This sounds like Modula-3 or Eiffel would be good languages for you.

   The call/cc and closures (though obviously elegant & useful) are IMHO
   not the primary reason for the existence and usability of scheme
   and other lisp clones. Witness early lisps.

If this was the case then we would still be stuck with those early lisps.

   The "If you need C you always know where to find it...'' manifesto
   might work perfectly for people who do not care much about efficiency,
   and for whom a 10-times (and more...) slowdown of their scheme procedures when
   compared to obvious C hacks is acceptable. There is nothing wrong
   with the attitude that speed is not so important; it is just that
   some people cannot afford to have that attitude, for various
   reasons. There are pretty different categories of users out there.
   Guess there is no need to attempt to satisfy everybody.

While it might sometimes be true (in my case) that speed doesn't
matter so much to me, this nevertheless shouldn't be overgeneralized.
It is very well possible to compile a languages with full
higher-order-ness and first-class continuations with a performance
comparable to C.  Witness Orbit, ChezScheme, SML/NJ, ...

--
-Matthias
