[plt-scheme] Extending

From: Matt Jadud (matt at jadud.com)
Date: Fri Aug 24 10:55:24 EDT 2007

On 8/24/07, Grant Rettke <grettke at acm.org> wrote:
> On 8/24/07, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
> > A similar effort at Harvard showed one thing: unless
> > students were told explicitly to use a debugger and
> > the repl, they wouldn't do it. Otherwise they couldn't
> I bet that resonates with most folks experience.

My initial work focused on the edit<->compile cycle, as with many
novices, there is a significant of (I argue) unproductive time spent
in this phase. Debuggers, test frameworks, and the like only become
part of their experience if they are able to write code that gets past
the compilation phase.

We have piloted the collection of the kind of data you're talking
about, Matthias, by looking at the spaces in-between successful
compilations (which, for many students, are rare), and examining how
they used the BlueJ environment during that time. I have nothing to
report on that, yet, as the data is new and I've recently crossed an
ocean, which discombobulates many things---especially my schedule and
workspace. Being able to capture some similar data re: students
following something like HtDP for the first time in Scheme would be a
nice addition to my explorations.

And while it is true that this may resonate with many people's
experience, I find the data is much richer than one might expect.
I've found the behavior of novice programmers, when examined closely,
to be a very interesting (and sometimes depressing, and sometimes
exciting) place.


Posted on the users mailing list.