Information Technology Reference
In-Depth Information
In the current paper, the reliance on rely conditions about equipment, rather
than a detailed description of the characteristics of particular equipment allows
for the replacement of the equipment, provided the new equipment meets the rely
conditions. On the other hand, monitoring of the healthiness of the equipment
may well (and probably should) be dependent on the detailed characteristics of
the particular equipment. By factoring out this aspect in the specification, the
specification can be more easily revised. A study of the contribution of other
research on “evolvability” to the issues of this paper will be undertaken in the
future. We wonder if there might be a way of using layers of rely conditions
where one set expresses things whose change would be disastrous while another
level is “anticipated evolutions”.
4.3
Further Developments
Our research contributes to the creation of specifications but it is informative to
look at how such specifications might be implemented. We know from sequential
programs that combining clauses of postconditions with
and
and
not
logical
operators provides a valuable way of recording “what” is required without saying
“how” it should be done. For example, the postcondition for a
Sort
routine can
be elegantly expressed as a conjunction of
InputPermutation
and
Ordered
.From
the discussion in Section 3.7 above, it looks as though one needs the full power of
a conventional programming language in order to “combine the machines” from
the various subproblems. One wonders whether new programming paradigms
could offer more natural “combinators” for such situations. (Another issue is
whether conventional programming languages like Ada or Java are ideal for
combining the sort of monitoring implied by the discussion in Section 3.6).
The research on “time bands” in [BB06,BHBF05] is extremely interesting
and we are already looking at ways in which time bands might help to achieve
a better structure for our specifications.
Another major avenue which we hope to pursue with our DIRC collabora-
tors Bloomfield, Littlewood and Strigini is handling stochastic assumptions and
requirements.
Acknowledgements
All three authors received support from the (UK) EPSRC funding of
Depend-
ability Interdisciplinary Research Collaboration (DIRC)
: the first listed author
was directly involved and the last two authors are Senior Visiting Fellows to
DIRC. In addition, the second author's research has been partially supported by
the Australian Research Council (ARC) Centre for Complex Systems, and the
first author's research has been partially supported by European IST RODIN
Project (IST 2004-511599). The first author now has funding from EPSRC under
the “TrAmS” Platform Grant and the EU's RODIN project.
We have derived great benefit from technical discussions with: Alan Burns,
Joey Coleman, Tom Maibaum and Jim Woodcock.
Search WWH ::
Custom Search