[geeks] just to stir things up, a few predictions

[Francisco Javier Mesa-Martinez]

On Sat, 23 Oct 2004 14:23:57 -0500 (CDT)
  "Jonathan C. Patschke" <jp at celestrion.net> wrote:
> On Sat, 23 Oct 2004, Charles Shannon Hendrix wrote:
> 
>>> I think you've got that backwards, since VMS is far 
>>>superior to UNIX.  :-)
>>
>> Yep, which is why Windows, based on VMS, is so 
>>incredibly good...
> 
> Windows is not based on VMS.  Windows is based on OS/2. 
> It just happens
> that Dave Cutler started working at Microsoft about the 
>time that the NT
> project started to come together, which is why NT has so 
>may VMSisms.


A friend who worked at M$ doing research, told me a 
different story. NT had nothing to do with anything while 
it was being "born." In fact there is a reason why the 
first NT to see the market was 3.x, there were previous 
releases of the system, at least interally, and these 
would seem pretty foreign to people who assume that NT is 
what came after 3.51. Basically it was a research project 
where Cutler and Co. were given free reign. It ended being 
a VMS-like  system using a microkernel and developed on a 
RISC machine with HW independence as a design methodology 
from the ground up, due to early experiences with the 
original target platform. VMS uses a monolithic kernel, 
and at least when NT was started it was quite tied to its 
native plaform (VAX)... so it was some sort of update on 
the ideas that cutler gor from working on systems like RSX 
and VMS. M$ also seemed to have hired an assload of the 
people working on Mach at CMU, the other half of the Mach 
group ended up at NeXT.

There was another project called OS/2 NT, it was killed by 
M$ when they split with IBM. So what we know as NT today 
is basically the marriage of the NT research, with the 
OS/2 NT, with the win32 cruft. Rumor has it that Cutler 
was beyond pissed at what was finally marketed as NT 
because it had very little to do with what he wanted, and 
a lot to do with what Bill wanted.

More trivia, NT is called that because the original target 
for the research project was the N-Ten processor (aka the 
80860). When it was clear that processor sucked big donkey 
balls and the x86 was thought to be a dead end (late 80's 
everyone was talking about RISC)... M$ and MIPS Computer 
developed together a reference platform for NT and named 
it the Magnum. In fact all the earlier development of NT 
was done on MIPS machines, and it seems that M$ had a 
large say on the design of the MIPS box. Which I think 
later was used as the reference platform for the ACE 
project. Who remembers those?

... man I feel old.