🚀 go-pugleaf

David Kastrup wrote in message ...
>"Larry Elmore" <ljelmore@montana.campus.mci.net> writes:
>
>> Dan DeClerck wrote in message <366D7988.A5B2EC9B@cig.mot.com>...
>> >Larry Elmore wrote:
>> >>
>> >> Pray tell, what technologies and innovations has Microsoft stifled?
>> >
>> >Intel's Native Signal Processing (NSP) which puts DSP like functions in
>> >the cores of CPU's.
>>
>> You mean it wasn't killed just because it's a dopey idea that isn't
needed
>> or even wanted?
>
>It was killed because it was an idea not wanted by Microsoft.
>Microsoft threatened not to support the MMX extensions of Intel in
>their OSs if Intel further pursued the project.  So Intel dropped it.
>At least that is the story told at the trials by Intel.

Which isn't the whole story, I'm sure. Intel's reputation isn't a whole lot
higher than Microsoft's in some respects. I'm not sure Microsoft really has
that kind of clout -- what would've happened if Intel had gone ahead with it
and told Bill Gates to go soak his head if he didn't like it? So what if MS
didn't support it? Everyone else would've supported it, leaving MS at a
competitive disadvantage. It's not as if they can suddenly say they're not
going to support Intel CPUs anymore...

>The motivation seems clear to me:  if Intel had continued, any
>operating system or software vendor could have fully supported all the
>MMX processing easily.  Now that Intel has had to drop NSP support,
>everybody has to implement his own code for using MMX and other
>stuff.  And that means that money does make a bit more of a difference
>for developers (having to develop more).

That makes sense, but I still don't see why Intel didn't tell MS to go jump
into a lake and implement NSP anyway, if it was so good. If MS didn't
support it, it would only hurt MS...

Larry