Path: utzoo!news-server.csri.toronto.edu!rutgers!usc!zaphod.mps.ohio-state.edu!casbah.acns.nwu.edu!eecs.nwu.edu!phil
From: phil@eecs.nwu.edu (William LeFebvre)
Newsgroups: sci.space.shuttle
Subject: Re: New (!?!?!?!) Shuttle Computers
Message-ID: <1991Mar11.201910.8476@casbah.acns.nwu.edu>
Date: 11 Mar 91 20:19:10 GMT
References: <1991Mar7.142311.10412@vaxa.strath.ac.uk> <6963@mace.cc.purdue.edu>
Sender: news@casbah.acns.nwu.edu (Mr. News)
Reply-To: phil@eecs.nwu.edu (William LeFebvre)
Organization: Northwestern University
Lines: 60
Nntp-Posting-Host: pex.eecs.nwu.edu

In article <6963@mace.cc.purdue.edu>, dil@mace.cc.purdue.edu (Perry G Ramsey) writes:

|> To be a little more fair, they have a very complex vehicle. It is
|> important to make sure that changes don't affect the system in unexpected
|> ways...

especially since any such "unexpected" behavior could cost us up to seven 
lives.

|> They've just gone overboard with the principle. 

That's a matter of opinion.

|> Working with outdated
|> equipment just because it was available off the shelf in 1974 ends up 
|> being a safety hazard, because you can't incorporate advances in capability.

I don't see that as a "safety hazard".  An impediment to progess, yes.
And perhaps an impediment to *increasing* the level of safety.

|> Besides the fact that NASA is supposed to be pushing technology if their 
|> existence is to have any value to the taxpayers.  The state of the Shuttle GPC
|> indicates pretty clearly that the current
|> NASA (at least the manned space side) is scared to death of anything new. 

If you want to see NEW technology being used at NASA, check out the 
UNmanned projects.  They aren't afraid to use new technology.  Why?
because the worst that can happen is we lose an expensive piece of
equipment.  If something goes mysteriously wrong with the shuttle, we
could lose LIVES!  The manned space program is intentionally safety
conscious.  They don't want any surprises.  Thus they stick with what
they know has worked in the past.  If they wanted to completely change
around the GPCs, using the most up-to-date technology that they think
would work in orbit (an important consideration, by the way), they would
have to start the orbiter's testing process all over again:  drop it
from a 727, have an initial test flight with two test pilots and with
ejection seats, etc., etc., etc.  By the time they were convinced that
everything worked smoothly, the technology would be woefully out of 
date.

Also, something people seem to have forgotten:  the GPCs are redundant.
They have to talk to each other to carry out this redundancy.  There's
some non-trivial hardware in there to make that happen.  You just can't
do that with 5 IBM PC clones clamped on an ethernet.  This goes far 
beyond the computers themselves.  All of that would have to be scrapped
and redone as well.  Why did they choose the 101-B in the first place?
Because they knew that it worked in fighters when they first started
designing this stuff.

|> That, though, is the ultimate state of any hidebound bureaucracy:  it's
|> better not to make a mistake than not to accomplish anything.

No, that is the ultimate state of anything that deals with human lives:
it's better to be safe and alive than to be cutting-edge and dead.

		William LeFebvre
		Computing Facilities Manager and Analyst
		Department of Electrical Engineering and Computer Science
		Northwestern University
		<phil@eecs.nwu.edu>