TITLE: The Program's the Thing
AUTHOR: Eugene Wallingford
DATE: February 04, 2008  6:51 AM
DESC: 
-----
BODY:
Readers of this blog know that programming is one of the
topics I most like to write about.  In recent months I've
had something of a "programming for everyone" theme, with
programming as a medium of expression, as a way to create
new forms, ideas, and solutions.  But programming is also
for computer scientists, being the primary mode for
communicating their ideas.

To the non-CS folks reading this, that may seem odd.  Isn't
CS about programming?  Most non-CS folks seem to take as
a given that computer science <EM>is</EM>, but these days
it is <EM>de rigeur</EM> for us in the discipline to talk
about "computing" and how much bigger it is than "just
programming".

Too some extent, I am guilty of this myself.  I often use
the term "computing" generically in this blog to refer to
the fuzzy union of computer science, software engineering,
applications, and related disciplines.  This term allows
me to talk about issues in their broadest context without
limiting my focus to any one sub-discipline.  But it also
lets me be fuzzy in my writing, by not requiring that I
commit.

Sometimes, that breadth can give the impression that I
think programming is but a small part of the discipline.
But most of my writing comes back to programming, and
when I teach a CS course, programming is always central.
When I teach introductory computer science, programming
is a way for us to explore ideas.  When I teach compilers,
it all comes down to the project.  My students learn
Programming Languages and Paradigms by writing code in
a new style and then using what they learn to explore
basic ideas about language in code.  When I taught AI
for eight or ten straight years, as many of the big
ideas as possible found their way into lab exercises.
Even when I taught one course that had no programming
component -- an amalgam of design, HCI, and professional
responsibility called Software Systems -- I had students
<EM>read</EM> code: simple implementations of
model-view-controller, written in Java, C++, Common Lisp,
or Ada!

I
<A HREF="http://www.cs.uni.edu/~wallingf/blog/archives/monthly/2006-12.html#e2006-12-02T10_42_40.htm">
   love to program</A>,
and I hope students leave my courses seeing that programming
is medium for expressing and creating ideas -- as well as
creating "business solutions", which will be the professional
focus for most of them.  Then again, the best business
solutions are themselves ideas that need to be discovered,
explored, and evolved.  Programming is the perfect medium
to do that.

So when I ran across
<A HREF="http://technically.us/code/x/the-nerd-factor-is-huge/">
   The Nerd Factor is Huge</A>,
via Chuck Hoffman at
<A HREF="http://nothinghappens.net/">
   Nothing Happens</A>,
I found myself to be part of the choir, shouting out "Amen,
brother!" every so often.  The article is a paean to
programming, in a blog "dedicated to the glory of software
programming".  It claims that programming needs an academic
home, a discipline focused on understanding it better and
teaching people how to do it.  (And that discipline should
<A HREF="http://www.cs.uni.edu/~wallingf/blog/archives/monthly/2007-10.html#e2007-10-04T18_40_13.htm">
   have its own conference</A>!)

In Yegge-like fashion, the author uses expressive prose to
makes big claims, controversial claims.  I agree with many
of them, and feel a little in harmony even with the ones
I wouldn't likely stake my professional reputation on.

<UL>
<LI> The shortage of women in computer science hurts our
     discipline, and it limits the opportunities available
     to the women, both intellectual and economic.  I
     would broaden this statement to include other
     underrepresented groups.

<LI> Trying to interest girls and other underrepresented
     groups by "expanding the non-programming ghettos of
     computer science" is misguided and insults these
     people.  We can certainly do more to communicate
     how becoming a computer scientist or programmer
     empowers a person to effect change and produce ideas
     that help people, but we should not remove from the
     equation the skill that makes all that possible.

<LI> Human-computer interaction doesn't belong in computer
     science.  Doing so "is as insulting to the study of
     interaction design ... as it is to computer science".
     This is perhaps the most controversial claim in this
     paper, but the authors makes clear that he is not
     showing to disrespect to HCI.  Instead, he wants due
     respect shown to it -- and programming, which tends
     to get lost in a computer science that seeks to be
     too many things.
</UL>

I agree with the central thesis of this article.  However,
separating programming as a discipline from HCI and some
of the other "non-programming ghettos" of CS creates a
challenge for university educators.  Most students come
to us not only for enlightenment and great ideas but
also for professional preparation.  With several distinct
disciplines involved, we need to help students put them
all together to prepare them to be well-rounded pros.

How should we encourage more kids -- girls and boys alike
-- to study computer science?  "Nerd Factor" is right:
don't shy away from programming; teach it, and sooner
rather than later.  Show them "how easy it is to create
something. Because that is what programming is all about:
making things."  And making things is the key.  Being able
to program gives anyone the power to turn
<STRONG>ideas</STRONG> into <STRONG>reality</STRONG>.

One of my early memories of OOPSLA comes from a panel
discussion.  I don't recall the reason for the panel, but
several big names were talking about what we should call
people who create software.  There were probably some folks
who supported the term "software engineer", because there
always are.  Kent Beck spoke heresy: "I'm just a programmer."
I hope I muttered a little "Amen, brother" under my breath
that day, too.
-----