Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!rex!wuarchive!julius.cs.uiuc.edu!zaphod.mps.ohio-state.edu!casbah.acns.nwu.edu!jln
From: jln@casbah.acns.nwu.edu (John Norstad)
Newsgroups: comp.sys.mac.programmer
Subject: Re: Gripes about System 7.0
Message-ID: <2936@casbah.acns.nwu.edu>
Date: 25 Jan 91 16:48:18 GMT
Sender: news@casbah.acns.nwu.edu
Organization: Northwestern University
Lines: 107
References:<5611@idunno.Princeton.EDU> <3810@uakari.primate.wisc.edu> <2899@casbah.acns.nwu.edu> <1991Jan24.224108.19413@ux1.cso.uiuc.edu>

In article <1991Jan24.224108.19413@ux1.cso.uiuc.edu> 
dorner@pequod.cso.uiuc.edu (Steve Dorner) writes:

> Bad, bad example.  I think Apple BLEW IT on AppleTalk.
> 
> The only interesting idea they had was the plug-and-play nature of the
> thing.  The protocol itself is horrendously myopic; it just won't scale 
to
> large internetworks.  That's why they've patched in Phase 2.

Hi Steve! 

In addition to plug-and-play (a MAJOR innovation), I would add NBP.

I don't find the protocol myopic at all.  It's just about as myopic as 
TCP/IP as far as I can tell.  In fact, they are very similar in many ways 
(e.g., RTMP almost = RIP, DDP almost = UDP, ADSP almost = TCP, etc.)

Everybody agrees that AppleTalk doesn't scale up well to large internets.  
It was unfortunately not designed with this possibility in mind - a major 
mistake.  Phase 2 does nothing significant to address this problem.  
There's an Internet Engineering Task Force working on the large internet 
problem.  For example, they're working on protocols so that separate 
AppleTalk internets can join together into a larger combined internet 
without having to worry about network number conflicts.

> I think Apple is catching up in networking, not leading the pack.

I would never call AppleTalk perfect, but it was indeed a major innovation 
and a tremendous success.  I'd say that Apple is helping to lead the pack.
 
> >IMHO, this is much more interesting, exciting, and important work than 
> >simply tacking "GUIs" on top of traditional command-line systems, e.g., 
X 
> >and NextStep on top of UNIX and Windows 3.0 on top of DOS.
> 
> I think you are missing something VERY fundamental.  The Macintosh is
> a GUI without an operating system.  Raw UNIX is an operating system 
without
> a GUI.  Modern systems will need *both*.

The Mac does indeed have an operating system - it has a file system, a 
memory management system, and a process scheduling system.  It was 
designed for personal computers rather than for traditional multiuser 
timesharing systems.  Yes, it's very different from those systems, and 
yes, it has alot of growing up to do, and yes, it could learn a number of 
very important lessons from it's big brothers.

The difference is that Apple has started over from the ground up.  They 
are rethinking traditional timesharing operating system concepts and 
redesigning them in very significant ways within the context of the direct 
manipulation graphical interface on a personal computer.  This takes a 
long time, and it will take many more years of hard work and lots of luck 
before we see the Mac OS, or more likely some successor to the Mac OS, 
become really mature.
 
> Apple is currently reinventing the operating system part; the UNIX 
vendors
> are trying their hands at the GUI.
> 
> Apple has a lot of problems in their task; making the Mac OS a real 
operating
> system (with VM, memory protection, preemptive multitasking, and a 
filesystem
> that doesn't drive you batty) isn't going to be easy.

Yes, this is all true, and in fact it may be impossible to extend the 
current Mac OS to include all this good stuff.  It seems likely to me that 
at some point in the not too distant future Apple is going to have to 
abandon the current OS and start over from scratch with a successor OS.
 
> The UNIX vendors are in a much better position.  There is nothing about
> UNIX which makes a good GUI hard to do; they have the freedom to solve
> the GUI problems correctly.  Some of the UNIX GUI's are dismal, laughable
> failures (eg, SunTools).  Others are arguably as good or better than the
> Macintosh (eg, NextStep).
> 
> There is nothing keeping these based-on-UNIX GUI's from being 
sophisticated
> GUI's with object-oriented message passing systems (that's what NextStep 
is).

Here's where we have a fundamental disagreement.  I don't think that the 
UNIX OS is a suitable platform for personal computers, no matter how many 
human interface "layers" you patch on top of it.  Personal computers in my 
opinion are radically different in a very fundamental sense from 
traditional multiuser timesharing systems, and they require new operating 
system concepts and designs, not just new GUIs on top of old operating 
system concepts and designs.

UNIX is the very best of the old tradition, and people involved in the 
design of new systems would be crazy not to study it carefully and learn 
from it.  But we need something completely new for the future of 
computing.  What I like about Apple is that they continue to have the guts 
to tackle this very difficult problem.

I agree, however, that NeXT is doing interesting things, and it's the only 
"GUI on top of UNIX" which I take seriously.

I wish I could figure out how I manage to get sucked up into these USENET 
arguments every few months.  I really should be doing some real work today 
:-) ...

John Norstad
Academic Computing and Network Services
Northwestern University
jln@casbah.acns.nwu.edu