Subject: Re: SGI ports [was Re: pwd_mkdb]
To: Jonathan Stone <jonathan@DSG.Stanford.EDU>
From: Harry W. Waddell <waddell@caravan.com>
List: current-users
Date: 04/09/1999 15:08:27
On Fri, 9 Apr 1999, Jonathan Stone wrote:

> 
> I've heard rumours of _interest_ from the direction of NASA Ames, but
> nothing more concrete.

I can speak for nasa ames, in an unofficial sort of way.

I think that "interest" sums it up pretty accurately. From time to time people
get pretty spun up about the notion of having one os running on all just about
every machine, especially after the vendors enter that recalcitrant phase,
e.g. when it dawns on them that no money will be going to them anytime soon.

I just wonder what the motivation would be? For someone running older SGI
stuff, which is about to get screwed by Y2k, and can't run IRIX 6.x, I
guess the motivation is clear. [I believe that certain "Challenge" class
machines fit this description]


 > 
> (Historically, SGI also tended to be very tightlipped about their
> graphics hardware, which'd be a big obstacle to getting graphics
> working; you may well want to emulate IRIX so as to run the SGI
> product X/GL servers, but that's just a guess).
> 

 I assume that most of you know about the snuggling up of SGI and linux. At
this point, they seem much more willing to just be a hardware company, except
for high end computing. One should be able to get information from SGI, if
they could show a real commitment to making their machines work with NetBSD.

Good emulation of Irix, with support for all the AV stuff would also be key
for any newer boxes, e.g. octanes, but there still may be good reason to do it
for the challenge machines. [Of course many of those are multiprocessor] So, I
guess we could save some of those older machines from the scrap heap. But, 
I really think that it would only be "interesting" for nasa for newer machines
since the processor power of older machines is pretty easily replaced at a
lower cost using more mundane things like i386 boxes. Hardware maintenance on
these old boxes is often enough to pay for replacement equipment.

I guess it would be another feather in Netbsd's cap to show that it can run on
yet another architecture, but I don't think that NASA is interested at this
time. NASA management would probably be just as happy if Linux ran on the
desktop machines. [for a variety of reasons that just make me morose. I'll be 
happy to discuss the reasons offline]

Harry Waddell
waddell@nas.nasa.gov