Xref: utzoo comp.lang.misc:7094 comp.object:2905
Newsgroups: comp.lang.misc,comp.object
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!think.com!snorkelwacker.mit.edu!stanford.edu!leland.Stanford.EDU!elaine13.Stanford.EDU!hoelzle
From: hoelzle@elaine13.Stanford.EDU (urs hoelzle)
Subject: Re: CHALLENGE: heterogeneous collections
Message-ID: <1991Mar27.042219.18924@leland.Stanford.EDU>
Sender: news@leland.Stanford.EDU (Mr News)
Organization: Stanford University - AIR
References: <1991Mar22.210725.29448@neon.S <1991Mar25.220525.11087@leland.Stanford.EDU> <26714:Mar2602:52:1891@kramden.acf.nyu.edu>
Date: Wed, 27 Mar 91 04:22:19 GMT
Lines: 27

> To implement a generic list type in C, you use macros. Nowhere is there
> an explicit ``insert'' operation that demands its own prototypes. So if
> you do have an operation that works on multiple types, it will work
> correctly through a list.
>
> Maybe one ``object-oriented'' language or another works badly with
> lists, but that's not related to static versus dynamic typing.

Really.   I'm not sure I understand your solution, so could you please 
post it?  Especially (if it is what I think it is), how does it generalize
to larger problems?  Just write more macros, I guess... 

Seriously: all programming languages are Turing-equivalent, so I'm sure
you can do everything in C.  But what about module interfaces (typed, of
course), abstraction etc?  How do you preserve them with macros?  How do 
handle non-toy examples where the equivalent to "insert" is a complicated
piece of code?  Do you really want an "inlined" copy of that code at every 
call site?

-Urs



--
------------------------------------------------------------------------------
Urs Hoelzle, CS PhD student			       hoelzle@cs.stanford.EDU
Center for Integrated Systems, CIS 42, Stanford University, Stanford, CA 94305