andrea laue wrote:
> In our final meeting of the semester, we will discuss a "straw man"
> curriculum proposed by Geoff and Worthy as well as the topics of
> digitization and sampling. Below you'll find the proposed
> curriculum.
I realized that some of the work John U. and I have done on the form of
the software engineering course has not been made available to the group.
I'm sorry we didn't make this available before now. You can find it on
the Web:
http://www.cs.virginia.edu/~horton/tmp/horton-unsworth-11-10.pdf
http://www.cs.virginia.edu/~horton/tmp/horton-unsworth-11-10-6up.pdf
These are copies of the Powerpoint slides that John presented at the
recent conference in British Columbia. The 2nd version of this is a 6-up
version of the slides. These reflect John's and my ideas, and have not
necessarily been seen or discussed by others.
I look forward to the discussion of the strawman. If the "software
engineering" course is taught in a way I'd prefer to see it taught, two of
the concepts related to requirements that I'd cover are "defining a system
boundary" and "requirements creep". I think we can see both issues here
in this proposal.
A system boundary is a precise definition of what a software system is
responsible for and what outside things (other systems, the operating
system, etc.) are responsible for and how they interact with the system
we're building.
One of the struggles I've had throughout our discussions are what's to be
included in which course, what kinds of skills will students have coming
into the software engineering course, etc. Am I wrong in perceiving that
the strawman makes some pretty dramatic changes to what's included in the
KR sequence? In particular the topics in Section 5 are now out of the KR
courses, yet many of these are the ones discussed in our seminar this term
or put on the schedule for next term. If it is a big change, then is it
one we all agree upon? Or do we need to go back to the beginning and
discuss the boundaries between the units in the curriculum.
"Requirements (or features) creep" -- you know this one! Let's continue
to add desirable things to what we're trying to build as we go along. The
strawman contains a lot of things that I'm in favor of -- don't get me
wrong! But there's now even more included in this two course sequence
than before, it seems to me. (Unless we're really not going to do any of
the topics in Section 5. But maybe still.) First, I think a lot of the
stuff that I had thought we'd do in the software engineering course has
been "moved down" into this first year sequence. (E.g. 3.8, testing,
maintenance and documentation.) I'd argue that in the first year they
look at big projects, do smaller individual kinds of things, and then in
the software engineering course they work in teams and migrate what
they've learned to a large-system, team-oriented situation.
I really would like to see programming taught during the first year
(somehow, somewhere). So don't get me wrong when I say that the technical
training components of this are pretty heavy, it seems to me. I count 26
days listed for Items 4.1-4.6, so let's figure that's a one- or two-hour
workshop once a week for two terms. So I'd guess these would be the "lab"
meetings for the KR course sequence? If so, does that preclude some other
kind of lab or discussion sections that others want?
There are a lot of things I like in the strawman, so sorry to focus on
just these. I look forward to the meeting tomorrow. (I might be late
since my exam ends at 11:30.)
Tom
P.S. One thing I don't like is the title "Management of Large Design" for
the SW Engin. course. (Of course! <grin>) I'll try to bring a list of all
the titles that have been suggested for us to look over tomorrow.
-- Dr. Tom Horton, Associate Professor Dept. of Computer Science, University of Virginia 151 Engineer's Way, P.O. Box 400740 Charlottesville, VA 22904-4740 Phone: 434 982-2217 FAX: 434 982-2214 horton@virginia.edu http://www.cs.virginia.edu/~horton
This archive was generated by hypermail 2b30 : Tue Dec 11 2001 - 12:32:52 EST