Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!rutgers!topaz.rutgers.edu!brandx.rutgers.edu!webber
From: webber@brandx.rutgers.edu (Webber)
Newsgroups: news.admin
Subject: Re: PD C as solution to binary groups
Message-ID: <315@brandx.rutgers.edu>
Date: Wed, 5-Aug-87 01:09:47 EDT
Article-I.D.: brandx.315
Posted: Wed Aug  5 01:09:47 1987
Date-Received: Fri, 7-Aug-87 02:49:19 EDT
References: <6960@g.ms.uky.edu> <3725@ncoast.UUCP> <2527@whuts.UUCP>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 36
Summary: be glad to help

In article <2527@whuts.UUCP>, tes@whuts.UUCP (STERKEL) writes:
> I do not know about the rest of the world, but lil-ol-me the
> run-of-the-mill user (non-hacker), has difficulty with about 70% of
> the C source put up on the Network.  At least 90% *does not compile

No need to act all bashful and such.  The question you are raising is
one that people who claim enough experience to moderate a sources
group have failed to appreciate.

> and/or link* leaving me to my poor wits to "fix".  Ultimately, about
> 70% remains worthless.  Most of this seems to be site/machine
> dependency of some sort, either by the poster or by my "home"
> machine.  I frankly do not see where a PD C would help at all.

The reason that you are having problems with all of these programs is
that they are usually sources that were only written under one version
of one C compiler on one version of Unix running on one particular
machine.  You would have no problem reconstructing the executable that
corresponds to the source if you had access to the C compiler,
operating system, and computer used by the person who created the
source.  What PD C does is make one C compiler accessible to all,
hence eliminating that weakness in the chain.  Because the language is
C, it is possible to write an entire operating system in the language,
thus eliminating a second weakness in the chain.  Unfortunately, there
isn't much that can be done about making sure that everyone one uses
the same computer.  However, the portability of C has been greatly
helped by the influence of the PDP-11/45 design on subsequent machine
architectures.

A completely different approach is to design a language that specifies
computations in a manner completely independant of the computer and
operating system on which it lies.  While some languages do this
better than other languages, no language does it well enough to tempt
me away from C (although I do use Lisp sometimes to prototype things).

---- BOB (webber@aramis.rutgers.edu ; rutgers!aramis.rutgers.edu!webber)