BRITISH COMPUTER SOCIETY FORTRAN SPECIALIST GROUP
Minutes of a meeting held on
Friday, 8 November 1974 at the
National Federation of Building
Trades Employers, 82 New Cavendish
Street, London W1 at 10.30 am.
Present: Mr B.H. Shearing (Chairman) Alcock, Shearing and Partners
Mrs E. Aylmer-Kelly University of York
Mr B.J. Banes Rolls-Royce, Bristol Engine Division
Mr E.O. Bodger IBM W.T. Systems Aid Centre
Mr P.D. Bond Philips Industries
Mr P.A. Clarke Rothamsted Experimental Station
Mr J. Cullen B.P.
Mr D. Ellison University of Manchester R.C.C.
Mr M. Fitzsimmons Berne Electronics
Mr D. Hill Macro Computer Systems Ltd
M I.D. Hill M.R.C.
Mr M.J. King B.B.C.
Mr C.K. Mackinnon U.K.A.E.A., Risley
Mr D.J. Maisey ICL
Mr K. Normington Lanchester Polytechnic, Coventry
Mr D.E. Oldfield ICL
Mrs R.J. Pyne Computing
Mr K. St Pier G.E.C.
Mr T.L. van Raalte Ministry of Defence, AWRE
Mr D. Winstanley University of Birmingham
Mr D.T. Muxworthy (Secretary) University of Edinburgh
Apologies Dr A.C. Day University College London
for Mrs F.E. Holberton ANSI X3J3/National Bureau of Standards
Absence: Mr T.D. Palmer Compower Ltd
Action
1. APPROVAL OF MINUTES The minutes of the meeting of 6 September 1974
OF PREVIOUS MEETING were approved subject to the correction of
typographical errors on page 5, section 2 line
5, 'permissible' and section 5 line 2,
'MAXREC=maxr'.
2. MATTERS ARISING FROM a. Mr Shearing said that there had not been
THE MINUTES a meeting Of the Specialist Group BHS
Chairman's Committee since the last
Fortran Group meeting but he had had
informal discussions about 'Computing'
It was thought that the recent
unanimous vote of confidence in
'Computing' by the Council did not
reflect feeling within the Society.
b. The Computer Journal Algorithms policy BHS
was also to be taken up at the next
meeting of the Specialist Group Chairman's
Committee. Mr I.D. Hill reported that
there were to be Algorithms Supplements
in the next two editions of the
Journal.
[page 2 of the minutes with agenda items 3 to 6 is missing - it should have been
reproduced on the back of page 1.]
Action
7. ANSI FORTRAN Decisions made at the July 1974 meeting of
ACTIVITIES Fortran subcommittee X3J3 (v. minutes of
6.9.74, Appendix A) were considered, as a
result of which comments were sent to X3J3
(v. Appendix A). Details of the September X3J3
meeting were not available (but see Appendix B).
There followed a detailed discussion of
part of chapter 12 of FORTREV, dealing with
input-output. This will continue at future
meetings: the OPEN, CLOSE and INOUIRE
statements remain to be considered.
Information on which systems allow mixed
formatted and unformatted records in a file
would be welcomed (cf. Second set of Fortran
clarifications, CACM Vol 14 (1971) pp.631-2,
topic 10).
8. STEERING COMMITTEE The Chairman asked for volunteers to form a
Steering Committee for the Group. This is
required by the ECS and the existence of an
active steering committee would enable the
Group to have more impact on the Society.
(After the Group meeting had closed the
following Steering Committee evolved:
Messrs Bodger, Clarke, Day, Ellison,
Mackinnon, Muxworthy and Shearing)
9. DATAFAIR 75 The Chairman pointed out that Datafair would be
held in London on 1-3 October 1975 and asked
for opinions on whether the Group should make a
presentation there.
10. PUBLICATION ERRORS There was some discussion on inaccuracies in
the article on Fortran by Peter Brown, published
in "Computing" on 31 October 1974. It was
decided not to reply formally as a Group but to
leave it to any individual who wished to reply
to do so.
11. NEXT MEETING The next meeting will be held on Friday, 24
January 1975 at 11.00 am; the venue is not yet
decided. Please note the change of time.
APPENDIX A DTM/PLU/1298
13 November 1974
To: X3J3
From: British Computer Society Fortran Specialist Group
Subject: Comments on items under discussion
The following points arose at our meeting on November 8, 1974. The most
recent X3J3 minutes available related to the July meeting so these comments
may be out of date.
1. We regret the rejection of the proposal that for the symbolic name
of an intrinsic function to be used for an external function, it must
appear in an EXTERNAL statement (/54,4.1). A proposal that an
explicit statement was preferable to a conflicting type statement was
approved (16-0).
2. On the equivalencing of dummy arguments (/54,4.9), three people thought
this was a good facility, eight thought it a bad facility.
3. (FORTREV 12.5.2). It was thought unclear what use could be made of
the ERR= exit in a list-directed READ. What is the effect of the next
READ?
4. (FORTREV 12.5.3d). What is an end of file in a storage file? Does
it imply array bound checking?
5. (FORTREV 12). There was some concern that the philosophy of defining
a Fortran file by the action of statements which write to the file might,
logically, lead to the situation where it was impossible to introduce
new data to a Fortran program. We did not have sufficient time to
consider this in detail.
SUMMARY OF DECISIONS TAKEN AT THE SEPTEMBER 1974 X3J3 MEETING
1. A proposal to clarify the side-effect of functions, when there is more than
one function reference in one statement, was passed (18-0).
2. A proposal to permit a character substring to be a storage file of one record
was passed (19-0).
3. Proposals defining subset levels 1 and 2 for sections 8, 11 and 13 of FORTREV
(respectively Specification, Control and Format statements) were approved.
4. A proposal to remove PRINT statements that specify a unit and WRITE statements
that do not specify a unit was passed (10-7).
5. a. A motion limiting the use for output of a free-field file being read was
passed (10-3).
b. A motion to prohibit the backspacinq of a free-field file was defeated
(7-9).
c. A motion to prohibit formatted input following free-field input from the
same file was defeated (3-13).
d. A motion to prohibit a file from containing both free-field and unformatted
records was passed (14-2).
6. A proposal to change keywords TYPE (in OPEN) and DISP (in CLOSED) to STATUS was
passed (16-1).
7. A proposal to remove < and > from the character set was approved (9-3).
8. A proposal to allow a processor to insert commas between free-field output
values was passed (9-1).
9. A proposal to allow restricted equivalencing of dummy arguments failed
(9-9).
10. a. A motion that free-field character output should not have delimiting
apostrophes was defeated (7-11).
b. A motion that all free-field character output should have delimiting
apostrophes, duplication of internal apostrophes and no carriage control
when the constant is longer than a line was defeated (2-16).
11. A proposal to allow free-field records in storage files was defeated (0-10).