Path: utzoo!news-server.csri.toronto.edu!rutgers!cmcl2!kramden.acf.nyu.edu!brnstnd
From: brnstnd@kramden.acf.nyu.edu (Dan Bernstein)
Newsgroups: comp.lang.misc
Subject: Re: Dynamic typing -- To Have and Have Not (was R
Message-ID: <25293:Mar1221:01:4491@kramden.acf.nyu.edu>
Date: 12 Mar 91 21:01:44 GMT
References: <608@optima.cs.arizona.edu>
Organization: IR
Lines: 23

In article <608@optima.cs.arizona.edu> gudeman@cs.arizona.edu (David Gudeman) writes:
> Alternatively: I fail to see how static typing can be regarded as more
> than a syntacitic feature.  If you're given a program that uses static
> typing, you can just check the types at every reference in the program
> to make sure they are correct...

That is correct. Since most people, most of the time, seem to want to
use statically typed values, and since compilers have an easier time
optimizing statically typed programs than dynamically typed programs, it
makes sense to provide static typing as the default. You can implement
dynamic typing on top of this if you think people will find it useful.

What is the argument here?

> BTW, you can't just change the program to refer to a (type,value)
> pair, you also have to insert code to do the right thing based on the
> type.

Didn't I say that? Yes, you have to do this, and it can represent a
significant amount of work (about as much as writing a C++ translator).
It's still a syntactic transformation.

---Dan