[Ti] Apple's True Market Share!

[Mark C. Langston]

On Fri, Dec 13, 2002 at 02:09:12AM -0500, Henry Kalir wrote:
> On Thu, 12 Dec 2002, Michael Bigley wrote:
> 
> > >This was a discussion about whether or not Apple should port to x86,
> > 
> > Actually it was a post about Apple's market share and this is an x-86 
> > tangent with no one changing the subject line...
> > -- 
> 
> Right! Lets take a vote here - How many of the listmemebers think that
> Apple SHOULD port OS X to the WINTELS?
> 


Personally, I think they'd benefit from the greater installed base of
hardware capable of running their OS.  I think users would benefit
from the concomitant increase in software ported to the OS.  However,
I think by expanding the diversity of hardware on which the OS must run,
the OS codebase would bloat and become more generalized, degrading 
performance on existing Apple kit.


So I'm ambivalent;  I was going to say that I feel the same way that
I feel about Sun -- that they need to realize they're a software, not
a hardware company, before they repeat the same mistakes SGI made.
However, Sun has gone a long way towards commoditizing and thus
marginalizing their hardware; they've removed a good deal of the
specialized engineering (S-Bus, lack of SCSI on lower-end models, etc.),
and they're left largely with commodity guts in a pretty candy shell
(the 10k and 15k excepted, and any other systems they design with
that type of dynamic reconfiguration in mind).  Going software-only
or spinning off the software division (or better yet, the hardware
division) entirely would probably be a good move all around.

Apple, however, still have tightly integrated hardware and software,
and an award-winning hardware design team.  Apple was originally a
hardware company, and continues to be so.  Rather than move away from
hardware -- something that would be inevitable, if they attempted
to compete in the commodity hardware arena -- they should insulate
the company from such a move, and create a daughter company which
would be responsible for the distribution and development of Marklar.
This way, if it's a failure, the parent isn't as hurt.  If it's a
success, the existing codebase is protected to a degree from the
subsequent bloat and generalization that selling it in-house would
require.

-- 
Mark C. Langston                                    Sr. Unix SysAdmin
mark at bitshift.org                                       mark at seti.org
Systems & Network Admin                                SETI Institute
http://bitshift.org                               http://www.seti.org