BRITISH COMPUTER SOCIETY FORTRAN SPECIALIST GROUP
Minutes of a meeting held on
Friday, 28 June 1974 at the
Royal Society of Tropical
Hygiene & Medicine, 28
Portland Place, London W1
at 10.30 am.
Present: Mr E.O. Bodger (Acting Chairman) IBM Systems Aid Centre
Mrs E. Aylmer-Kelly University of York
Mr B.J. Banes Rolls Royce (1971) Ltd
Mr P_D. Bond Philips Industries
Mr H.W. Bradly CAD Centre, Cambridge
Mr F. Chambers Logica Ltd
Mr P.A. Clarke Rothamsted Experimental Station
Mr J. Cullen B.P.
Dr A.C. Day University College, London
D. Ellison Manchester R.C.C.
Mr M. Fitzsimmons Beme Electronics
Mr I.D. Hill M.R.C.
Mr J.P. Holland Tunnel Cement Ltd
Mr D.J. Maisey I.C.L.
Mr W.Y. Milne Independent Computer Consultations
Mr D.E. Oldfield I.C.L.
Miss M. Spooner Computing
Mr T.L. van Raalte A.W.R.E.
Mr D. Winstanley University of Birmingham
Mr D.T. Muxworthy (Secretary) University of Edinburgh
Apologies Mr E.H. Shearing (Chairman) Alcock Shearing & Partners
for Absence: Mr J.S. Gatehouse I.C.L.
Mr T.D. Palmer Compower Ltd
Mr K. St Pier G.E.C.
ACTION
1. ACTING CHAIRMAN Mr Bodger chaired the meeting as Mr Shearing
was indisposed.
2. APPROVAL OF MINUTES OF The minutes of the meeting of May 3, 1974
PREVIOUS MEETING were approved.
3. MATTERS ARISING FROM The BCS Standards Committee had indicated
THE MINUTES that it was willing to underwrite the
reproduction in Britain of the draft
American Standard for Fortran to the extent
of E75. It had suggested that only 40
copies be circulated for comment and that
these remain the property of the BCS;
further that a master copy be held in the
ECS library from which would-be purchasers
could obtain a photocopy.
The meeting thought that most of the active
Group members would wish to purchase a copy
and that they would be prepared to pay, say, ALL
£1.50 to help finance the printing. It was
decided that those members who wished to buy
their own copies, and those who wished to
be put on the circulation list for BCS
copies should inform the Secretary of this
(the address is now: Program Library Unit,
18 Buccleuch Place, Edinburgh EH8 9LN;
please do not send any money yet).
It was pointed out that official ANSI
approval had not yet been obtained.
4. NEW BCS RULES FOR Mr Bodger summarized a new set of rules for
SPECIALIST GROUPS Specialist Groups which were being drafted;
these had been discussed at the most recent
Specialist Group Chairmen's meeting and
substantial changes had been suggested.
The general intention of the rules was to
make the Specialist Groups more active.
One rule which was likely to be adopted was
that each group should have a committee of
at least three people of which at least one
should be a member (small m) of the Society.
The meeting felt that in the rules no
distinction should be made between members
and Members and that more responsibility
should be given to groups, in that relevant
matters related to standards, refereeing,
etc should automatically be referred to
them.
Copies of the next draft will be available
from Messrs Bodger or Muxworthy. Mr Bodger
again asked any group member who wished to All
serve as vice-chairman of the group to make
himself known to the officers; this applies
also to the office of secretary.
5. SPECIALIST GROUPS' Mr Bodger had represented the Fortran Group
CHAIRMEN'S MEETING at the last Group Chairmen's meeting: he
reported that a new meetings room was being
built at BCS HQ and that the Specialist
Groups were to have an overhead projector
provided especially for them. Other items
reported included the facts that Eurcomp
made a £5,000 profit for the Society, and
that there had been little progress with
the Society's Science Museum project.
The new series of the Computer Bulletin,
the first edition of which was due out in
September 1974, was expected to have 40
pages per issue, of which 8 would be
advertisements.
6. ANSI X3J3 MAY MEETING The minutes of the May meeting of the
ANSI Fortran Committee, X3J3, were not to
hand but Mr Maisey had received a letter
giving details of the main decisions
reached, viz.
1. A motion to require that the index of
an implicit DO be defined at an END=
or ERR= exit was defeated 9-11.
2. A motion to prohibit leading zeros on
output when using the Iw code was
approved 20-0.
3. A motion that blanks produced on
output by the Gw.d code be true
spaces and not skipped characters was
approved.
4. A motion to allow conversion of a
constant in a DATA statement to be of
the same type as the corresponding
variable in the variable list, pro-
vided this was of type integer, real
or double precision was approved 20-1.
5. A motion to allow the comma between a
constant list and a variable list in
the DATA statement to be optional was
approved 16-5.
6. A motion to restrict the DO-loop
control variable to be of type
integer was defeated 7-14. (This was
the third time at recent meetings this
motion had been voted on, always with
the same decision).
Mr Maisey also reported that certification
of various sections of FORTREV were
scheduled for the next few X3J3 meetings
and that certification of the complete
document was due at the January 1975
meeting. The most recent FORTREV was
dated 1 May 1974 and a new edition was
expected within a few weeks.
7. ANSI X3J3 INPUT-OUTPUT There was a discussion of some of the input-
PROPOSALS output proposals in FORTREV. some members
of the group were unhappy about the defini-
tion of fi1e-names and of direct-access
files, and about the fact that NAMELIST did
not appear in FORTREV. A memorandum was
sent to X3J3 (see Appendix B).
8. OTHER BUSINESS a. A Group member, Mr D.T. Hall had written
to suggest that the Standard should
include a requirement that compilers be
able to produce listings with line
numbers, cross-reference lists of
variable names and labels, warning
messages for certain logical errors and
so on.
It was pointed out that the Standard
used the term "processor" rather than
"compiler" and although the meeting
agreed with the sentiments of Mr Hall's
letter, previous experience in this area
(of the availability of n in STOP n, and
the difficulties of the Groups's own
Diagnostics working Party) indicated
that there was no point in pursuing this
matter at the moment.
b. The Society had asked the Group for its
Annual Report; this is shown in
9. DATE OF NEXT MEETING The next meeting will be held on Friday 6
September 1974 at 10.30 am in the new
Meeting Room, ECS Headquarters, 29 Portland
Place, London W1.
[No record of the meeting of 6 September 1974 has survived. Click here for the
next meeting (8 November 1974) for which records are available.]
BCS FORTRAN SPECIALIST GROUP ANNUAL REPORT 1973-74
1. The main activity of the Group has continued to be cooperation with the
Fortran committee, X3J3, of the American National Standards Institute
in their drafting of a new standard for Fortran. The draft is expected
to be open for public review early in 1975 and the Group has been making
plans to facilitate the review in the U.K.
2. The Group has established a liaison with the Fortran committee, TC8, of
the European Computer Manufacturers' Association and it has closely
followed the progress of the British Standards Institute Programming
Language Committee, DPE/13, particularly so far as Fortran is concerned,
for no fewer than four active Group members are also members of DPE/13.
3. An offer by the Group to check algorithms published in the Computer
Journal for compliance with the Fortran Standard has been accepted.
4. The Group has continued to be a forum for the discussion of many
matters related to Fortran and it has also taken a positive interest in
relevant areas of the Society's affairs.
5. There have been five meetings during the year, with an average
attendance of 18; there are about 110 people on the Group's mailing
list. The Chairman during the year was B.H. Shearing and the
Secretary was D.T. Muxworthy.
D.T. Muxworthy
To: X3J3
From: British Computer Society Fortran Specialist Group
Subject: Comments on current proposals
The following matters arose from discussion at the Group meeting on
June 28, 1974.
1. Leading zeros on output
We have not yet had an opportunity to study the minutes of the X3J3
May meeting but feel that there is a need to be able optionally to
print leading zeros, for example in sums of money on cheques and in
navigational courses. Our proposal for an M format code (/42,G)
would allow this.
2. Filename (12.2.2)
We are still unhappy that the class of filenames may be null on some
processors. This means that a standard-conforming program using a
filename will not necessarily be transportable and hence the
filename facility is of dubious value.