Time Magazine: Man of the Millennium

George Herbert gherbert at crl5.crl.com
Sat Sep 26 14:26:38 EST 1998


Peter da Silva <peter at baileynm.com> wrote:
>Craig Burley  <burley at tweedledumb.cygnus.com> wrote:
>>  Linus Torvalds
>>At least from a comp.arch point of view, he's probably done more for
>>the viability of "whatever architecture you want to create and
>>whatever machine you want to build, as long as it's cost-effective"
>>as anybody.
>
>Now now, it's just the fluke of a lawsuit that held BSD back long enough
>for Linux to take the lead, and I suspect that without Linux and BSD the
>Hurd would have been FORCED to switch from "development toy" to product
>by now. That's three free operating system projects peaking in the same
>few years. And that's just the UNIX-like ones... there's been a plethora
>of specialized open-source operating systems.
>I wouldn't give primacy to Linus, though he's an amazing fellow... but
>rather to Dennis Ritchie, Ken Thompson, Brian Kernighan... [...]
>And right under them, Andy Tannenbaum... because without Minix [...]

Entirely too software-oriented from the comp.arch point of view.
Though it predates my professional activities, it appears from the
histories and texts that the revolution in Unix and C was more
of a particularly good, internally consistent and scaleable
implimentation of ideas that were around and in general use.

I'm not quite sure which single person to credit it to (Moore?),
but the rise of the microprocessor from early LSI seems to be the
technical path that enabled the tremendous consistent growth we've seen.
Everything else could have taken an alternate path and we would
still have possibly made it here... PCs for a long time didn't
use Unix or C, and even now the dominant desktop OSes have only
borrowed UNIX concepts though C and relatives are the development
standard today.  The Internet started on non-UNIX systems and could
have continued to grow on them.   But without widely affordable
microprocessors of rapidly increasing capabilities,
computers would have remained unaccessable.  Software has always
muddled through, giving in relatively short order reasonable
capabilities scaled to whatever hardware is available.
The hardware has to be there for that to mean much,
though, and microprocessors were what enabled us to have
the hardware here, there, and everywhere.


-george william herbert
gherbert at crl.com




More information about the Neur-sci mailing list