Subject: [gcproposal-0] Suggesting a Roadmap
Date: Thu, 16 Oct 2003 17:07:58 +0100
From: Robin Milner <Robin.Milner@cl.cam.ac.uk>
X-Scanned-By: MIMEDefang 2.29 (www . roaringpenguin . com / mimedefang)

Dear all

In the current phase of the GC Exercise, each GC proposal group is
considering its next steps. These typically take the form of a network
or workshops aiming to refine the proposal in the direction of a
platform for a Grand Challenge.  We offered some concrete suggestions
to the Moderators on how this refinement might be done; it seems a
good idea also to air these suggestions in the GC0 discussion group.

[AS: that's here:
    http://archives.nesc.ac.uk/gcproposal-0/
]

We suggest that each GC proposal group should aim to produce a series
of supporting documents.  These will have considerable value in
promoting the quality and effectiveness of UK Computing Research, even
if the eventual conclusion is that there are less risky ways of
achieving the research objectives than by committing to an
international Grand Challenge project.

A crucial intermediate supporting document could well be a 'Roadmap of
strategic directions'.  It should state the long-term scientific goals
of the research, in terms of extending scientific knowledge and
understanding and engineering skill.  It should survey the known
effective research methods in all areas contributing to the goal.  It
should identify the pathways through which the results from one
research area may be passed on for the benefit of another.  In some
cases, it might include a list of open research questions to which the
answers are within reach.

The Roadmap will have independent value as a reference for subsequent
grant proposals, which need not then repeat justification for their
research goals, at least not where they coincide with those of the
Roadmap (without discouraging new directions in any way). This kind of
contribution to drafting of research proposals was one of the intended
roles of the original 'Foresight' documents; but our GC Exercise is
motivated primarily by advancement of scientific ideals rather than by
any immediate potential for commercial exploitation.

The necessary firm basis in the underlying science can be assured by
another useful document: a comprehensive Survey of the current state
of the art.  This should identify the leading researchers and teams
worldwide. It might include an annotated bibliography, background
reading for anyone entering the field as a new researcher.  Indeed, it
is hard to imagine a convincing Roadmap without such a Survey.

A Popular Science Presentation of the scientific goals of the proposal
would also be valuable.  We suggest that it should appeal primarily to
University entrants, who can be inspired by ideals of scientific
curiosity to choose the most relevant areas for their own study and
research.  It should also appeal, perhaps through journalists, to the
general public and hence even to politicians. In a popular science
presentation, one might confess to a little science fiction about the
way in which the knowledge and understanding gained by scientific
research could eventually be exploited.

We are also much concerned with the international dimension.  Perhaps
the time is now right at which a few international leaders could be
contacted about their general views, and the prospects of mounting a
collaborative initiative in their own country.  Perhaps we could
suggest suitable topics for a Dagstuhl meeting, or for a Newton
Institute Seminar, or a Royal Society Discussion Meeting.  It is worth
bearing in mind that these -- at least the last two -- have to be
proposed some three years before they occur.

It would be well still to avoid couching any of our documents as an
implicit appeal for additional funding.  The message should be that
that it is the scientist's job to say how funds can be most
efficiently spent in pursuit of their scientific objectives.

A final point of this message is to reinforce confidence that the work
devoted to setting out a Roadmap, in awareness of the possibility of a
GC project, is an excellent way to promote long-term research.  There
is a good prospect that, in whatever way is most appropriate, the
contributions made to the GC Exercise will play a decisive role in
advancing Computer Science.

Tony Hoare, Robin Milner