On behalf of the DoCS academics I'd like to present a proposal for a new
subject to the Teaching Committee. (I don't expect it to be discussed
at tomorrow's meeting, but it should be tabled in time to allow safe
passage through the authorities in time for being on the books next
year.)

The proposal form can be found at
http://www.cse.unsw.EDU.AU/~gernot/Stuff/proposal.{tex,ps}. The
motivation and further information can be found in the mail attached
below.

Gernot
- --

Reply-to: G.Heiser@unsw.edu.au
Date:    Thu, 04 Mar 1999 12:01:01 +1100
To:      docs.pg
cc:      docs.as
From:    Gernot Heiser <gernot@cse.unsw.EDU.AU>
Subject: Proposed new Subject: Readings in Computer Systems

Dear students,

I circulated the below proposal among DoCS academics a few weeks ago. As
the feedback was positive I'm now widening the discussion to include
you.

The suggestion has been made (and I support this) that the student
participants would be involved in assessing their peers. This is based
on the experience that students are quite able to make reasonable
assessments of their peers' performance, and that it is good practice
for them. This does NOT mean backing away from only allocating a
satisfactory/unsatisfactory grade for the subject, but would most likely
have the form of a "mock mark" given to students as part of the feedback
on their contributions. [ Note that this subject is meant to be for
research students, and they only get a satisfactory/unsatisfactory grade
for their thesis, so introducing marks into subjects only meant for
research students would serve no purpose at all. ]

What do you think? The subject, if it goes ahead, would be on the books
for next year, but we could actually run it this year, because it
doesn't take much to set up.

Gernot

- -----

From: gernot@cse.unsw.edu.au
To: docs.as
Subject: Proposed new Subject: Readings in Computer Systems
Content-Type: text/plain; charset=US-ASCII
Date: Thu, 18 Feb 1999 16:24:57 +1100
Sender: gernot@zuse.disy.cse.unsw.EDU.AU

Colleagues,

stumbling across a few WWW pages I noticed one significant shortcoming
in our programs compared to the top US places: we don't have any
subjects were you systematically and in-depth look at hot research
papers. I was considering setting up something along these lines for OS,
but then thought that my students would benefit to some wider exposure
to research trends, e.g. in computer architecture, networks,
databases. Hence the proposal for a subject covering the wider area of
computer systems.

The idea would be to have a subject (which we'd make mandatory for
research students, as part of their 45 credit point requirements) where
each student gets allocated one paper, or a small number of related
ones, which they analyse thoroughly and then give a half hour seminar on
them. (There could be a written report as well, but I'm not sure about
this.) This would be followed by a discussion of the issues presented in
the paper. All students enrolled would be required to attend all
seminars, and each would give one of them. At the end they get a
"Satisfactory" grade (except in the unlikely case where they are deemed
a failure). Undergrads would not normally be allowed to enrol (although
there may be exceptions, particularly for final year double degree
students who have already been involved in research projects) but would
be encouraged to attend the seminars.

Each enrolled student would be assigned to an academic working in the
field of their paper for guidance. As the academic would be quite
familiar with the material, the supervision effort would be quite
small.


What do you think about this? If people like the idea in principle, I'd
widen the discussion to involve research students.

Gernot