17 Jan
2014
17 Jan
'14
8:06 p.m.
On 01/17/2014 01:56 PM, Al Kossow wrote:
On 1/17/14 9:34 AM, John Wilson wrote:
One has to remember what the machines were used for. It was designed to
solve very large
problems that create huge arrays. Oil mapping is one thing that does
that, the others explode.
So likely the reason is not so much cray as the specific application or
project it was used for.
Hence the close lipped people.
It was my understanding that a smaller machine managed it and fed it and
accepted
fire hose rate data from it and there was likely no OS but a framework
that the specific
application was supported by. It was from all I read not used as a
general purpose machine
like a VAX or CDC 6000. So rather than OS you likely looking for
something like VAX ELN,
PDP-11 IOX, or some other library of "stuff".
Allison
he'd brought a copy home back in
the old days and still had the tape, so he was wondering how to
handle this
obvious good news w/o getting in big trouble for what was technically
theft.
I'll bet there was a lot of that! Especially if you count hardcopy.
Yes, though you had a very small set of clients for a Cray, and they
weren't
very happy with people just taking stuff out of the building.
I spent some time digging for Cray software when I first started at
CHM, since
we have machines and some documentation but no software at all. I
found out
pretty quickly that people who worked on the machines really didn't
want you
to ask them about that.
There is a large class of big machines where this is true. Hell, it's
even true
once you get away from the popular minis. Try finding supermini or
array processor
software from the 80's today.