Newsgroups: comp.software-eng,comp.lang.functional,comp.lang.lisp,msu.cps.misc
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!news.sprintlink.net!tank.news.pipex.net!pipex!uknet!newsfeed.ed.ac.uk!edcogsci!jeff
From: jeff@cogsci.ed.ac.uk (Jeff Dalton)
Subject: Re: Why typing?
Message-ID: <DFH6K2.4qG@cogsci.ed.ac.uk>
Organization: Centre for Cognitive Science, Edinburgh, UK
References: <43rcpn$26re@info4.rus.uni-stuttgart.de>  <19950921T152258Z@naggum.no> <BLUME.95Sep21153818@atomic.cs.princeton.edu> <DFBCKn.7os@aplcenmp.apl.jhu.edu>
Date: Mon, 25 Sep 1995 19:09:37 GMT
Lines: 29
Xref: glinda.oz.cs.cmu.edu comp.software-eng:37323 comp.lang.functional:6464 comp.lang.lisp:19209

hall@aplcenmp.apl.jhu.edu (Marty Hall) writes:

>In article <BLUME.95Sep21153818@atomic.cs.princeton.edu> 
>blume@atomic.cs.princeton.edu (Matthias Blume) writes:
>[...]
>>When I still was a Scheme addict I used to believe in this, too.  But
>>after programming extensively in SML I now have to say that this is
>>not true at all.  Early _complete_ error checking leads (at least for
>>me) to a huge decrease in development time.

>I have been and continue to do a majority of my professional
>development work in Lisp (the remainder in C), and like the dynamic
>typing aspect. But for those who haven't tried SML, it really is worth
>looking at. I have considered the strongly typed aspect of other
>strongly typed languages a pain, but SML [...]

I agree that SML is a good language and that ML-style type
inference, and the power of the type system, make for a usaable
and flexible form of strong-typing.

However, I nonetheless find SML a pain to use.  I don't think
it's the strong-typing _per se_, though I still find it too
restrictive.  I hate having to write output routines (Lisp
typically provides enough automatically), pattern-matching
is often too "positional", ML's not object-oriented, it
doesn't have macros, the module system is in itself about
as complex as all of Scheme, ... -- stuff like that.

-- jeff
