🚀 go-pugleaf

John,
 
>>I'd appreciate comments from people who have programmed for a variety of
platforms.
 
OK, I think I qualify here, having spent 12 years with Unix and a variety of
minis, mainframes and PCs before coming to the Mac about six years ago (and not
exclusively since then).  And John, I agree with you.  One can admire a
twelve-foot high scale model of the Statue of Liberty as a remarkable
engineering and artistic accomplishment and still question whether the artist
chose a suitable medium.  I have to pinch myself here.  Am I missing something,
or is at least part of the logic of this thread going as follows: (A) the Mac
is a great platform, (B) the Mac doesn't have a modern OS, (C) one can, through
heroic efforts, achieve high quality anyway, therefore (D) a modern OS is a BAD
THING?  Let me put it this way: were a protected-mode, truly multitasking,
modern operating system available for the Mac, why would one NOT want to use
it?  Outside the PC segment, not having one is unthinkable.  Why?  Because it
is widely available and has substantial value.
 
It seems that much of the DOS/Windows and Mac communities have drawn the wrong
lesson from the personal computer revolution of the 80's.  The PC (both kinds)
was successful not because it contained more primitive operating systems and
tools but because it addressed problems that were not being addressed at all
otherwise and these economies led people to produce less reliable but
low-priced and very profitable products.  The environment has changed.  Quality
expectations are up, both because of shifting patterns of usage and competitive
pressures.  Complexity is up.  Machines are being used for a variety of
purposes rather than being purchased for a single application.  Businesses,
even big ones, are using networks of PCs as the backbone of operations,
replacing or shoving into the background the minis and mainframes that used to
provide the reliable anchors.  Many of them are finding their way into
mission-critical usage, such as clinical laboratories and aircraft design, even
on-board systems.  The PC architectures of yesterday just are not up to these
standards of software quality.  It's like using bottle caps when bronze is
needed: Yes, people produce good results anyway, but is that a positive
statement?
 
At the same time, one can look at the trend line in a very important conceptual
ratio and see that it is dramatically dropping as applications get ever more
complex: overhead resources (CPU cycles, disk space & etc.) consumed by a
"real" OS divided by resources consumed by applications.  And it is a much
talked-about problem in the computer community that hardware (e.g., resources)
are getting faster and cheaper at a much faster rate than software usage
expands to use the added resources.  Put another way, the relative performance
and resource cost of a real OS is dropping like a rock.
 
In short, it is time to stop being so provincial and start learning from people
who have been addressing these sorts of problems for many decades elsewhere in
the computing community.  We don't have the same constraints we had in the past
so let's act accordingly.  This is certainly a large part of what has driven OS
efforts by IBM in OS/2, Microsoft in Windows NT, and Apple both internally and
through Taligent.  It is also what has driven Sun, HP, Silicon Graphics and
others from the Unix world into the PC arena.
 
In development tools, the PCs are perhaps leading rather than following.  That
I find sad because they are still primitive, no matter how one tries to justify
them.  We can do much better.
 
Regards,
Jeff Alger
SBM International