[Ti] Apple's True Market Share!

[Mark C. Langston]

On Thu, Dec 12, 2002 at 11:08:57PM -0600, Chris Olson wrote:
> Ric Perrott wrote:
> 
> > Intel was around and making x86 processors long before Microsoft was 
> > anything.
> 
> Sure they were, but you're trying to say that Intel would be where they 
> are today if Microsoft had never existed?  Intel designs their stuff to 
> run Windows, and they work very closely with Microsoft to make sure that 
> new designs are compatible with new versions of Windows.  Intel doesn't 
> give a hoot about linux or anything else that runs on x86.


Oh, come now.  Are you claiming that the x86 instruction set has been
optimized for an operating system?  That's like claiming that LEGOs
are optimized for a particular model.

Yes, Intel and Microsoft talk quite a bit.  I'm sure Intel gives
Microsoft early access to architectural changes and instruction-set
additions, changes, and deletions.  And I'm sure Microsoft asks Intel
to ensure certain opcodes are more efficient, simply because their
compiled code makes heavy use of a few repeatedly.  But again, you're
making claims with no support -- this leaves you looking suspiciously
like you've got an axe to grind rather than a point to make.  You've
already made it clear that you go out of your way to avoid Microsoft
and Intel.  Are you certain your opinions aren't coloring your
ability to make objective assertions?

Surely there are enough underhanded things that both companies have
done that are well-documented that you don't need to resort to
hypothesizing more?

Windows runs on AMD processors as well.  Are you also claiming
AMD's optimizing their opcodes for Windows?


I guess Intel was optimizing their opcodes for Commodore prior, while
Motorola were opimizing theirs for...who?  Commodore?  TI? 
Apple?

I'd love to meet the guy at Microsoft who can look at the millions
of lines of sourcecode that make up their product line and determine
what atomic operations need to be most efficient, and what new 
atomic operations would improve the overall efficiency.

I haven't met him the last few times I've been in Microsoft's HQ or
Silicon Valley satellite campus.  None of my acquaintences at 
Microsoft know him.  Perhaps it's all smoke and mirrors and
optimizing compilers and working to cut the bloat just enough to
assure continued _minimum_ performance on new architectures instead?
Nah.  Sounds too plausible.  But hey, I'm guessing.  I admit it.
You?



-- 
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