Hi JohnK,

I think that you may well be right. Don't know that it would ever get to multi-CPUs on a single workstation (though I'm seeing MBs advertised now that claim to DUAL a Celeron) because "splitting" that OS along the lines you suggest might still be problematic for reliability. But I never guessed an Internet access through a microwave oven either, so who knows??

I see a possibility, though, of MS making a very plain-Jane non-GUI flavour of NT with reliability and performance as the major design factors.

But it may yet be, as you suggest, that Linux will have an impact on this, even in the near future. I suppose that if MS cannot find ways to listen to its users then they will have to 'read' their users by watching some of them move to Linux.

I sure hope that something happens to wake MS up on this. Way way back, in the early 70s, IBM systems regularly required more than 1 IPL per day. Once we had live users attached (TSO or online systems) this simply was unacceptable. It really didn't take them that long to straighten it out, but then we had mechanisms then to let them know, in no uncertain terms and in large volume, that it was a problem in need of immediate attention.

Let's keep hoping.

Cheers,

Jim N

>Hiya JimN.
>
>On a purely philosophical note, I wonder if the pendulum is swinging back the other way on UI vs. power/reliability? Look at the rising popularity of (mainly) non-GUI Linux. Linux kicks NT's butt in speed and, in most studies, reliability....but no one would ever accuse it of being end-user friendly.
>
>It makes me wonder; is it possible that future desktops will be designed with multiple CPUs, one to handle user interactions with the big, fat, friendly GUI and another backend CPU(s) running a much more simplistic OS divorced from the GUI to do the brute processing? Sort of an internal client/server architecture. Now THAT would be an intriguing architecture IMHO.