On Mon, Mar 19, 2007 at 12:50:43PM -0700, Mike Cheponis wrote:
> http://www.kroah.com/log/linux/ols_2006_keynote.html claims:
> 
> "Linux supports more different processors than any other operating system 
> ever has."
> 
> "Yes, we passed the NetBSD people a few years ago in the number of 
> different processor families and types that we support now."
> 
> Is that true?  And has it been true since 2004?  ("a few years ago")

Possibly, in the sense that Linux is a trademark used by very different
kernels... for two reasons:

a) Linux marketing guys count things like uCLinux, which runs on
   processors without memory management units. We'll probably never
   support this, because virtual memory is integral to a lot of
   kernel structures (e.g.  buffer cache) and, given that even
   specialized embedded CPUs come often with an MMU nowadays, nobody
   feels enough pain to decouple this.

b) Linux marketing guys tend to count processors that can't (yet) be
   built out of the central source tree - because, at least
   historically (don't know the status quo), the central source
   tree isn't accessible to people != Mr Torvalds. We normally don't
   do that.

c) You always have to be very careful when you compare what's different
   and what not - is m68k one family/cpu type? two (010 and >=020)? three
   (010, >=020, 020/30 with Sun MMU)? PowerPC - hashed vs. tree MMU? etc.

However, they do support some CPUs (S390) that were sponsored by the
vendor as a technology demonstration, and that we don't.

Regards,
	-is
-- 
seal your e-mail: http://www.gnupg.org/