In working out the present report on the Conference on Software Engineering
organised by the NATO Science Committee, every attempt was made to make it
useful to a wide circle of readers.
Thus parts of it are written for those who have no special interest in
computers and their software as such, but who are concerned with the impact of
these tools on other parts of society.
This class includes, for example:
·
civil servants
·
politicians
·
policy makers of public and
private enterprises.
These readers should find material of interest in Section 1 (Background of
Conference) and Section 2 (Software Engineering and Society).
A somewhat narrower readership for the report includes those who need an
understanding of the nature of software engineering, although they are not
themselves working in the field. These
readers are typically:
· managers of business enterprises using computers
· researchers in fields other than software engineering and computer science
· university officials
· computer marketing personnel.
These readers should find points of interest in Section 3 (Software
Engineering), Section 7.1 (Software: the State of the Art), Section 7.2
(Education), and Section 7.3 (Software Pricing) as well as in Sections 1 and 2.
Finally, a large part of the report addresses itself to those directly engaged
in the design, production (implementation), and service of software. These technical areas are first given an
approximately uniform coverage in Sections 4 (Design), 5 (Production), and 6
(Service). The succeeding chapters 7
(Special Topics), 8 (Invited Addresses) and 9 (Working Papers), present more
detailed treatment of a selected set of topics.
The main problem in deciding on a structure for the report was to decide
between one of two different basic classifications, the one following from the
normal sequence of steps in the development of a software product, from project
start, through design, production or development, to distribution and
maintenance, the other related to aspects like communication, documentation,
management, programming techniques, data structures, hardware considerations,
and the like. The final structure is
based on the first of these two classifications. However, in many places an unavoidable influence from the second
type of classification has crept in.
The editors are only too aware of this problem and have attempted to
mitigate its effects by provision of a detailed index.
The text of the report derives mainly from two sources, viz. the working papers
contributed by the participants before or during the conference (mostly in June
1968), and the discussions during the conference. The discussions were recorded by several reporters and most were
also recorded on magnetic tape. The
reporters' notes were then collated, correlated with footage numbers on the
magnetic tape, and typed. Owing to the
high quality of the reporters' notes it was then, in general, possible to avoid
extensive amounts of tape transcription, except where the accuracy of
quotations required verification.
However, to give an impression of the editors’ task, here is an example,
albeit extreme, of the typed notes:-
536 DIJKSTRA
F -
H --
P --?--
(here ‘536’ is the tape footage number,
and the letters F,H and P identify the reporters).
This section of tape was transcribed to reveal that what was actually said was:
» There is tremendous difference if maintenance means
adaptation to a changing problem, or just correcting blunders. It was the first kind of maintenance I was
talking about.
You may be right in blaming users for asking for blue-sky equipment, but if the
manufacturing community offers this with a serious face, then I can only say
that the whole business is based on one big fraud. [Laughter and applause] «.
For use in the report the source texts, and some additional
transcribed material, have been sorted out according to the subject
classification developed during the conference. Whenever possible the material in the working papers has been
integrated into Sections 3 to 7 of the report.
However, in some eases it has been found more convenient to place the
working material in Section 9, and merely to make the appropriate references in
the main sections of the report.
To avoid misinterpretations of the report it must be kept in mind that the participants at the conference were acting as individuals, and in no sense as representatives of the organizations with which they are affiliated.
In order to retain the spirit and liveliness of the conference, every attempt
has been made to reproduce the points made during the discussion by using the
original wording. This means that
points of major disagreement have been left wide open, and that no attempt has
been made to arrive at a consensus or majority view. This is also the reason why the names of participants have been
given throughout the report.
The actual work on the report was a joint undertaking by several people. The large amounts of typing and other office chores, both during the conference and for a period thereafter, were done by Miss Doris Angermeyer, Miss Enid Austin, Miss Petra Dandler, Mrs. Dagmar Hanisch, and Miss Erika Stief. During the conference notes were taken by Larry Flanigan, Ian Hugo and Manfred Paul. Ian Hugo also operated the tape recorder. The reviewing and sorting of the passages from the written contributions and the discussions was done by Larry Flanigan, Bernard Galler, David Gries, Ian Hugo, Peter Naur, Brian Randell and Gerd Sapper. The final write-up was done by Peter Naur and Brian Randell, assisted by Ian Hugo. The preparation of the final typed copy of the report was done by Miss Kirsten Andersen at Regnecentralen, Copenhagen, under the direction of Peter Naur.
Peter Naur
Brian Randell
Update 30-JUL-2004