I went to an Engineering for Complex Systems workshop at Pittsburgh,
October 21,22. Day One
was a bit of waste of time
but
Day Two
ended in an uncharacteristic burst of optimism about the future of
software engineering at NASA.
|
Attendees |
The usual suspects were there.
Ken McGill: a
man who truly loves his cookies.
Martha Wetherholt- a woman with her fingers on the pulse
and ears to the ground
(shown her fingering
the pulse on her ears).
Larry Markosian from ARC (my Strat3 colleague),
John Kelly;
Martin Feather; and
Sally Godfrey.
Also present were various interesting folks from around NASA.
and the Software Engineering Institute (SEI) guys who facilitated it all.
I want to meet the architect for the SEI building- SEI's front door
reflects the church across the road. This is no accident.
SEI also reflects
a religious belief by government agencies
in the benefits of SE process.
Ah, architecture. Gags that last generations.
|
|
Day One: Ungood |
Day one was spend made the usual trite statements about how
NASA's software problems could be greatly reduced if we only had the
funding to apply basic software stuff (inspections, requirements
tracability matrices, etc). But NASA negotiates its dollars from
Congress and Congress never gives us what we want so we make do with
less in year one, hoping that in year N we can back load the
work.
Meanwhile, the planets don't care. When they present their
launch windows, we usually have to blast off- even if we aren't 100%
ready (that's for unmanned deep space
missions- manned space fight and earth orbit stuff
is another story all together).
|
|
River Cruise |
Monday evening, we went on a river cruise.
To hear the locals talk, we were all going to swept away.
I wasn't looking forward
to it but even Pittsburgh
couldn't make
twinkling lights reflected in the water look anything else than magical.
It was real pretty and it was a good chance
to talk to everyone.
|
|
Day Two: Good News |
Day two, Ken and me nearly left early. Ken actually challenged the SEI
guys "tell me why I should stay". That shook things up a bit and
the agenda changed, for the better.
Tim Crumbley (from Marshal) showed
us the a number of Data Request Documents (DRDs) which he had proposed to
the next phase of the Orbital Space plane (OSP) development. Tim's
DRDs were amazing. If adapted, they would likely provide a software
manager with the detailed data needed to identify software problems in
time to do something about them. AND it would mean researchers have real
data to work on.
Then we discussed strategies for effecting change. The room came to
realize that many of the things we want changed had already been
started in the last 2 years. For example, the DRDs which Tim Crumbley
presented were actually written by a NASA software working group
member for the last two years. The Subgroup hasn't been terribly
successful getting NASA to agree on or adopt a software metrics
program, but the thinking and in many cases, exact words from subgroup
document were obvious in Tim's DRDs.
And other attitudes emerged in Day Two that were most encouraging.
There were several examples where belt way people publicly and
vehemently defended the IV&V Facility. At one point Ken made the
statement that on certain projects the IV&V analysts had more insight
into the project than anyone in the government. Tim Crumbley
challenged the statement. Immediately Martha jumped up an defended
it, by naming projects where she felt the Titan IV&V analysts had more
insight than the government people managing the development. Talk
about a cultural change!