9 Apr
2000
9 Apr
'00
9:13 a.m.
Richard Erlacher said,
The conventional wisdom where NASA is concerned, is
that they HAD high
standards through the Apollo program and that shortly thereafter, a lot of
people left and apparently took vital talents with them. I've not worked
directly with NASA people in a very long time, and can't agree or disagree
with that view.
I'm working at NASA right this very minute. I think it's pretty much
impossible to categorize "NASA" as a single monolithic coding entity. Here
at GSFC, there's one group running the ground system that is very
conscientious and writes good code, with input checks, comments, version
control, descriptive variable names, etc., and properly tests the code
before running it for real. Then there's another group where their
acceptance test run of the code (which I observed) gave a list of internal
tests, all saying "passed". But when I asked exactly what each test did, it
turned out that *half* of them did nothing other than print up the word
"passed". This was flight code, btw, and they had not told us of any plans
to upgrade it. It's better now, and fortunately is not critical to the
mission in any case. But the bottom line is that even within a single NASA
field center, there's a *lot* of variation.
- Mark
9 Apr
9 Apr
9:55 a.m.
New subject: NASA programming (was: Re: !Re: Nuke Redmond!)
----- Original Message -----
From: Mark Tapley <mtapley(a)swri.edu>
To: <classiccmp(a)classiccmp.org>
Sent: Sunday, April 09, 2000 3:13 PM
Subject: OT: NASA programming (was: Re: !Re: Nuke Redmond!)
it
Richard Erlacher said,
>The conventional wisdom where NASA is concerned, is that they HAD high
>standards through the Apollo program and that shortly thereafter, a lot
of
>people left and apparently took vital talents with
them. I've not worked
>directly with NASA people in a very long time, and can't agree or
disagree
with that view.
I'm working at NASA right this very minute. I think it's pretty much
impossible to categorize "NASA" as a single monolithic coding entity. Here
at GSFC, there's one group running the ground system that is very
conscientious and writes good code, with input checks, comments, version
control, descriptive variable names, etc., and properly tests the code
before running it for real. Then there's another group where their
acceptance test run of the code (which I observed) gave a list of internal
tests, all saying "passed". But when I asked exactly what each test did,
turned out that *half* of them did nothing other than
print up the word
"passed". This was flight code, btw, and they had not told us of any plans
to upgrade it. It's better now, and fortunately is not critical to the
mission in any case. But the bottom line is that even within a single NASA
field center, there's a *lot* of variation.
- Mark
It's hard to come up with generalized remarks that will apply to every part
of the organization. However, the attitude that pervades the top mangement
trickles down considerably. I was not referring to coding standards alone.
It's likely that there were few enough tests in the system you described
that someone wanted it looking more like one out of a large number of tests
was failing, when really half of the two relevant and valid tests were
failing.
Perhaps my duty on the "shuttle-return-to-flight" project turned me off on
these practices, but I've learned that NASA goes to great lengths to put on
the appearance of being thrifty. They are, of course, thrifty when matters
of human life are all that's concerned, but when it's about their continued
funding, they'll spend a lot just to make smoke. When we turned in our
report, our mangement wanted us to rewrite it so it would pass. The
managers had to send it out without the signature of any of the engineers,
however, because of their practice of ground-ruling-out some of the most
dangerous aspects of the shuttle's operating procedures. I understand that
most of the '86 hardware has been updated. Perhaps someone with a
conscience had a part in that and corrected some of the system flaws.
Dick
8975
days inactive
8975
days old
1 comments
2 participants
participants (2)
-
edick＠idcomm.com -
mtapley＠swri.edu