eugene@pioneer.UUCP (03/30/87)
You asked for comments.
Let's face it. Computer science is not a science (it's aspiring). I
should also note these is a discussion group on the ARPAnet of soft-eng
which has been quiet, too bad some of your (whole net) comments are
good ones. (Send mail to mday@mit-xx.arpa)
The first time I encountered the fact it was not a science (I'll never
forget this) was when I was a freshman in a dorm commons. Simon (a
Senior EE) and I (I was Nuc. Eng. at the time) had a discussion that
it's not really a science, but rather a technology or engineering.
Science (Simon asserted) was the domain of the natural world. [Herbert
Simon's "Sciences of the Artificial" had not yet been written.]
Computer science up until recently was basically a spinoff of
mathematics and philosophy. Both hold a separate place when compared to
more classical sciences like physics and chemistry. I think what
distinguishes science from say philosophy is increased emphasis in 1)
improving observation, 2) experimentation and hypothesis testing.
It can be argued (By Denning at lunchtime) that different sciences have
different criteria for judging work.
I think what makes a science is the tension between the theoretical and
the experimental. And what will improve computer science is more
experimental work (real experimental work). BTW: proof is not
experimentation, nor is simple testing equal to experimentation.
This all relates to software engineering because SE has such a poor base
(foundation) and will require more. You will see greater effect of
software engineering when the principles trickle into Freshman and lower
division classes rather than relegated to upper Div. classes. When the
intro to progamming class uses a high-level design tool, then SE has
made progress. At best it can be argued that "structured programming"
has at least made inroads "down there" (when structured programming was
argued as software engineering).
From the Rock of Ages Home for Retired Hackers:
--eugene miya
NASA Ames Research Center
eugene@ames-aurora.ARPA
"You trust the `reply' command with all those different mailers out there?"
"Send mail, avoid follow-ups. If enough, I'll summarize."
{hplabs,hao,ihnp4,decwrl,allegra,tektronix,menlo70}!ames!aurora!eugene