[2762] in WWW Security List Archive

home help back first fref pref prev next nref lref last post

Re: UNIX in Perspective (was: Re: DOS and Macro Virus Discussion)

daemon@ATHENA.MIT.EDU (Fisher Mark)
Thu Aug 22 15:56:55 1996

From: Fisher Mark <FisherM@is3.indy.tce.com>
To: hallam <hallam@ai.mit.edu>, www-security <www-security@ns2.rutgers.edu>
Date: Thu, 22 Aug 96 11:02:00 PDT
Errors-To: owner-www-security@ns2.rutgers.edu


>The point I was trying to make was simply that the UNIX vendors put
>no resources into fixing problems with UNIX that could have been fixed
>for a tiny proportion of the profits it made for AT&T.
Yes, and no one else is, either (commercially).

>I don't think that the claim that WNT is the successor to UNIX holds
>for a moment. Cutler's contempt for C is apparent in the compiler he
>wrote. WNT is very clearly the successor to Culter's earlier systems,
>ELN and VMS.
Here at TCE, our common hardware platforms are IBM PCs, and Sun and HP 
workstations.  The commercially available OSes in common release for these 
platforms are:
* DOS/Windows -- The less said, the better.
* UNIX -- As we have already discussed.
* Linux -- Still based on the underlying UNIX paradigms of a monolithic 
kernel and security (apart from limited file security) as an afterthought.
* Windows NT -- Which at least does security for OS objects correctly, and 
is a microkernel so should be easier to maintain and update.  The problem 
here is that higher-level network objects (Web pages et. al.) are not OS 
objects, so the security and reliability must be applied after the fact. 
 Unfortunately, Cutler appears to like to do things the hard way (as you 
have noted, his problem with using high-level languages).  In this way he 
may resemble Gates, as I have seen much more elegant designs for IBM PC OSes 
than MS-DOS and the Windows kernel.  It almost seems Microsoft's fate is to 
always go down the most difficult path towards a technical goal.

>Its all very well for programmers to think they design systems for
>programmers. Really what they are doing is designing systems for
>experts. There is little that is technically innovative about the
>Web. What made it a success was that it was designed from a users
>perspective and not from that of a technocrat looking to increase his
>job security through obscurantism.
Computers are still too hard to use, whether you are an expert or not.  An 
interesting article, if you haven't seen it, is "The Anti-Mac Interface" by 
Don Gentner and Jakob Nielsen in the August 1996 Communication of the ACM. 
 I especially noted how the use of language could help tame the complexity 
of what we want to use computers for.  I don't think (for the most part) 
that UNIX or most OSes have their enhancements designed just for 
obscurantism; instead, if the trade-off is between "increased power" and 
"usability by non-experts", the designer will almost always opt for 
increased power.  "Do not attribute to malice what can be adequately 
explained by stupidity."

Just my 0.02$US...
======================================================================
Mark Leighton Fisher                   Thomson Consumer Electronics
fisherm@indy.tce.com                   Indianapolis, IN

home help back first fref pref prev next nref lref last post