British Computer Society Fortran Specialist Group

Minutes of a meeting held on 5th April, 1976 in the Staff Common Room,

at the Polytechnic of Central London, 115 New Cavendish Street, London W1

Present:

Mr. D. T. Muxworthy (Chairman)    University of Edinburgh.

Mr. M. J. Appleford               Ove Arup and Partners.

Dr. J. C. Baldwin                 University of Cardiff.

Mr. M. Black                      Philips Industries

Mr. P. D. Bond                    Philips Industries

Mr. M. Dolbear                    B. P.

Mr. J. L. Dyke                    Huntingdon Research Centre

Dr. P. L. Evans                   Polytechnic of Central London

Mr. R. M. Greenwood               M.R.C.

Mr. G. L. Harding                 Leeds University

Mr. W. S. Hilder                  M.R.C.

Mr. D. Hill                       M.C.S.L.

Dr. I. D. Hill                    M.R.C

Mr. D. J. Holmes                  Rolls Royce (1971) Ltd.

Dr. J. S. Hutton                  S.R.C. (Rutherford Laboratory)

Mr. M. J. King                    B.B.C.

Mr. M. R. Lewis                   Imperial College Computer Centre

Mr. I. McNicol                    ESA/ESOC Darmstadt

Dr. J. Murchland                  UCL

Mr. K. Normington                 Lanchester Polytechnic

Mr. T. L. van Raalte              M.O.D.

Mr. R. W. S. Rodwell              ICL

Mr. M. Smith                      Rolls Royce (1971) Ltd.

Mr. C. Tampin                     P.M.A.

Mr. P. Vyse                       Honeywell I. S.

Mr. P. A. Clarke (Secretary)      Rothamsted Experimental Station.

Apologies for Absence:

Dr. A. C. Day                     UCL.

Mr. D. Ellison                    UMRCC.

Mr. B. Shearing                   Alcock, Shearing and Partners.

1.   Approval of Minutes.

The minutes of the meeting of 2nd February, 1976 were approved subject

to the following corrections:

Attendance: Mr. W. S. Hilder, MRC was also present at the meeting.

Item 1: "organiser of datafair" should read "Chairman of specialist

groups committee".

2.   Matters arising from the Minutes.

Item 2(c) The £60 requested from the BCS for completion of this year's

programme of meeting has been obtained.

Item 2(c) Subscriptions (payable by non-members of the BCS) for minutes of

meetings. After discussion with the BCS it has been decided to charge

£1-00 to cover costs of distributing minutes to non-members of the BCS.

One member had enquired if institutional members would be required to pay.

The BCS said that there was no precedent for this and they would raise it at

the next Specialist Groups Meeting. Currently the BCS have been allowing

three members from each institution to attend conferences at the normal BCS

member charge. In cases where more than three members attend, the excess have

been charged at non-member rate. In this light, as a provisional measure,

the Fortran Specialist Group will allow three individuals from each institution

free of charge for minutes of meetings.

The position of tax relief on subscriptions is being investigated.

3.   Election of Officers for 1976

Two resignations from the steering committee have been received. Colin

MacKinnon of UKAEA Risley has withdrawn because of pressure of work. Eric

Bodger of IBM has now moved to the United States until 1978.

Those present wished to thank Colin and Eric for their services to the group

and in particular to Eric who as vice-chairman had been the group's

representative on the BCS Specialist Groups committee, had chaired several

meetings and given his support to make the Datafair 73 seminar on Fortran a

success.

Both Chairman and Secretary asked if there was anyone willing to take their own

places, but as no names were forthcoming, were re-elected for a second year.

Requests for someone to take the vacant post of vice-chairman were not

successful and this will be reconsidered at the next meeting.

Requests for new members of the steering committee resulted in Messrs.

G. L. Harding, D. Hill, M.R. Lewis and J. Murchland joining the existing

members A. C. Day, H. D. Ellison and B. H. Shearing as the new steering

committee. Later Mr. Shearing suggested that we ask Eric Bodger to act as a

special liaison with the U.S.A.

Eric had expressed the possibility of helping the group co obtain information

from any appropriate meetings which take place in California.

4.   Revision of objectives

Following further discussion, the wording of the proposed revised objectives

now becomes "To undertake activities associated with any aspects of Fortran".

It is intended to present this for approval at the next Specialist Groups

meeting.

5.   Activities of other Fortran Groups

5.1  ANSI X3J3

Since the last meeting the draft proposed standard had been published. The

50 copies purchased by the Fortran Specialist group had all been ordered, and

orders are still arriving. It is possible that some orders may be cancelled

and if so, the copies will be made available to those on the "waiting" list.

Mr. P. Vyse had written an article on the draft standard which had appeared

as a three part serial in "Computing" newspaper. Mr. Muxworthy thanked

Mr. Vyse for writing the article and complimented him on the way he had

presented the subject.

The minutes of the December meeting of X3J3 were described briefly by Mr.

Muxworthy and it was noted that meetings were scheduled for 10th February,

25th April and 11th July 1976.

A discussion followed and Mr. Normington raised some objections to the

PARAMETER statement. He thought that the syntax would be more regular if the

list of parameter definitions was enclosed in parentheses. He objected to the

parameter definitions because the type of the parameters is determined by their
usage. This is irregular within Fortran. He received some measure of support

from those present on both these points. He also thought that the restriction

on names to six characters should be removed to- permit more-meaningful names.

Mr. McNicol said that he was disappointed that symbolic "length" attributes for

variables (other than type CHARACTER) had not been included in the new

standard. He pointed out that ICL had introduced such a feature in the ICL 2900

series Fortran, which allowed variables of several lengths.

There was some discussion on the best way of getting public comment on the

draft standard. The possibility of collecting comments, summarising them and

sending a 'concensus' view to X3J3 was rejected on the grounds that no one was

willing to do the summarising and that some comment would be lost in the

process.

There was also a problem with timing. Our next meeting is on the 7th June 1976

and any concensus reached at the meeting would have to be drafted and sent to

X3J3 before the 30th June deadline, which it might not make.

It was decided that in order to ensure that comments were timely the best

approach would be for individuals to send their comments directly to X3J3

and also to send a copy to a "collector" for the group to discuss in June

on the basis of which, it is hoped to form a "group view" on the standard

which we could send to X3J3 to re-inforce the individual views.

Comments for X3J3 should be sent to:                ACTION: All members

Lloyd W. Campbell,

X3J3 Secretary,

BRL-CSD, Bldg. 328

Aberdeen Proving Ground, MD2l0O5

USA.

and a copy for the Fortran Specialist Group's collector to:

Mr. M. Lewis,

Imperial College Computer Centre,

Room 401, RSM,

Imperial College,

Prince Consort Road, LONDON SW7

Dr. I. D. Hill made the point that positive comments indicating approval

of certain items were as important as criticisms and would help to form

a more balanced view.

For those who do not have copy of the draft standard, several summary documents

are available. These include Mr. Vyse's article in computing (ll, 18 and

25 March 1976), a summary by W. Brainerd in the Fortran Development Committee's

FOR-WORD newsletter No. 5, Oct. 1975 (copies available on request) and a

review by Dr. L. Meissner entitled "Proposed ANS X3-9 Fortran Language
Revision" which occupies the entire contents of FOR-WORD Vol. 1 No. 6 Jan 1976

and is currently being circulated to members.

5.2  Fortran Development Committee activities

Since the last meeting we have received copies of FOR-WORD Vol. l No. 6

Jan. 76 which contains a summary of the "Proposed ANS X3.9 Fortran Language

Revision" and Vol. 2. No. l. Feb. 76 contains an updated list of preprocessors,

reviews and correspondence.

5.3  Codasyl Fortran DBML Committee Activities

The minutes of meeting number 10 held on 5th Jan. 76 have been received.

The FDBMLC Journal of development, version 0.4 was revised by inclusion

of a Glossary of terms and a problem associated with the ERASE statement

and currency indicators which prevents access to the next record was uncovered.

6.  Progress Reports

6.l FORTREV review Working Party

Mr. Muxworthy reported for Dr. Day that the Working Party had completed its

Examination of FORTREV and a copy of this had been sent to the Computer

Journal for inclusion in the May 76 issue. The latest information was that

the article would not appear in the May issue because of delays. However,

this was not as disastrous as first thought because the article by Frank

Engel, Chairman of X3J3 had also not been included in the May issue. It is

believed that both may now appear in the August 76 issue.

6.2 Proposed Promotion Working Party

Mr. Clarke outlined the objectives and scope of activities of the proposed

Promotion Working Party. The objective is to collect and distribute information

about the BCS Fortran Specialist Group. This would take the form of information

for use within the BCS, information for new and prospective members of the

group and once set up would involve routine updating. Much of this information

has already been collected as part of the Secretary's normal duties. A request

for persons to join this working party failed to get any response.

6.3 Preprocessor Requirements Study

Dr. Murchland as convener outlined some areas of possible study for this

working party (see Appendix B). The word 'preprocessor' was not exactly

apt and the intention was more that Fortran Text should be examined to find

the ways in which it could usefully be manipulated. One approach was to

study some of the existing processes that manipulate Fortran Text and a

provisional list was read out (see Appendix B). A request was made for

persons to join this working party. Messrs. P.A. Clarke, P. L. Evans

and I. McNicol volunteered to join Dr. Murchland on this study. Anyone

else wishing to join this working party should contact Dr. Murchland directly

at:

Traffic Studies Group,

University College London,

Gower Street,

London. WClE 6BT

7.  Any Other Business

Mr. Clarke reported that two copies of the draft proposed ANS Fortran standard

had been placed in the BCS library and were available for loan. The BCS Library

had also kindly given the group a list of all books relating to Fortran in

the library together with details for making loans (see Appendix A). The group

will be informed of new acquisitions relating to Fortran.

Mr. Muxworthy said that the E.C.I. (European Co-operation in Informatics) were

arranging a conference from 9-12 August 1976 on the Advancement of Informatics

and that they were also setting up specialist groups.

He reported that the BCS had published a report on the "Beaumont Weekend",

at which many leading members of the society had discussed the topic of the

"Future of the BCS".

8.  Date of Next Meeting

The next meeting will be held on 7th June 1976.

9.  Talk on DEC Fortran Systems

Dr. D. Clifford described the features of the Fortran Systems available on DEC

PDP8 and PDP11 computers (see Appendix C). This was followed by questions and

discussion. A photocopy of the slides presented is available on request from

the Secretary.

APPENDIX A

Borrowing Regulations of British Computer Society Library

Members may borrow up to four books at any one time.

Books are issued for an initial loan period of one month.

Books may be renewed after this original period if they are not required

by another reader.

Books which have been renewed after the original loan period may be recalled

if required by another reader.

Books which are already on loan may be reserved and the requested will be

notified when the book is available.

A charge of 40p per item is made for postal loans to cover the cost of postage

and packing. Cheques/postal orders made payable to the Librarian, City University.

Books of 5 prepaid vouchers ( 2.00 per book) are available from the library

for members who intend to use the service regularly. Members bear the cost

of returning books by post.

Photocopies of journal articles can be supplied, from journals in stock,

at a charge of 5P per page, subject to the provisions of the Copyright Act.

The library acts as a source of information on all aspects of computers and

their applications. Lists of books and periodical articles on well-defined subjects

can be supplied including details of items which are not in the library's stock.

Library hours: During University term  09.00-2l.00 Monday-Thursday

09.00-20.00 Friday 09,30-12.30 Saturday morn.

During Univ. vacation   09.00-l7.00 Monday-Friday

I enclose details of the books we have in stock on Fortran,

together with regulations for borrowing.

A number of the books will be out at any one time, particularly

the new ones, and I advise members to ring the library

before calling in to borrow a particular item and we will

check the shelves.

We are always interested in receiving suggestions from members

for additions to stock.

Items are usually only available for loan to members of the

Society. l believe that members of Specialist Groups are not

necessarily members of the Society. it will therefore be necessary

for non-members to borrow through members.

Yours sincerely,

(signed) Judy Thomas

J.M. Thomas (Mrs.)

BOOKS ON FORTRAN IN BCS LIBRARY

COLMAN, H.C,: Computer language: an auto instructional introduction to Fortran.

McGraw Hill, 1962.

FORSYTHE, A.I. et al: Computer science: Fortran language. Wiley, 1970.

COURT, R.: Fortran for beginners. Holmes McDougall, 1970.

BAJPAI, A.C. et a1: Fortran & Algol: a programmed course for students of

science and technology. Wiley, 1972.

HAAG, J.N.: Comprehensive Fortran programming. Hayden, 1965.

HAAG, J.N.: Comprehensive standard Fortran programming. Hayden , 1969.

HARRIS, L.D.: Fortran programming: (2 & 4) Merrill, 1964.

BEZNER, H.C.: The elements of Fortran IV. Goodyear Pub., 1973.

CALDERBANK, V.J.: A course of programming in Fortran IV. Chapman & Hall, 1969.

DOCK, V.T.: Fortran IV programming. Reston, 1972.

FRIEDMANN, J. et al: Fortran IV: for self study or classroom use. Wiley, 1975.

HEALY, J.J. & DEBRUZZI, D.J.: Basic Fortran IV programming: self instructional

manual and text. Addison Wesley, 1968.

HEALY, J.J. & DEBRUZZI, D.J.: Basic Fortran IV programming. Addison Wesley, 1975.

JAMES, E,, O'BRIEN, F. 8. WHITEHEAD, P,: A Fortran programming course. Prentice

Hall, 1970.

JAMISON, R.V.: Fortran programming. McGraw Hill, 1966.

KALLIN, S.: Introduction to Fortran. Auerbach, 1971.

KAZMIER, L.J. & PHILIPPAKIS, A.S.: Fundamentals of EDP and Fortran: a self e

instructional manual. McGraw Hill, ]970.

KENNEDY, M. & SOLOMON, M.B.: Tan Statement Fortran, plus Fortran IV: for the

IBM 360 featuring the WATFOR and WATFIV compilers. Prentice Hall, 1970.

KUNZI, H,P,: Numerical methods of mathematical optimization with Algal and Fortran

programs. Academic Press, 1968.

LOUDEN, R.K.: Programming the IBM 1130 and 1800. Prentice Hall, 1967.

MCCAMERON, F.A.: Fortran logic: and programming. Irwin, 1968.

MCCAMMON, M.: Understanding Fortran. Crowell, 1969.

MCCRACKEN, D.D.: A guide to Fortrnn IV programming. 2nd edition. Wiley, l972.

MCCRACKEN, D.D.: A simplified guide to Fortran programming. Wiley, 1974.

MCCUEN, R.H.: Fortran programming for civil engineering Preentice Hall, 1975.

MALCOM, R,E.: Computers in business: a Fortran introduction. International

Textbook Co., 1968.

MARXER, E.: Elements Of computer programming: Fortran. Delmar, 1973.

MEISSNER, L.P.: Rudiments of Fortran. Addison Wesley, 1971.

MURRAY-SHELLEY, R.: Teach yourself computer programming: Fortran. English UP, 1967.

NATIONAL COMPUTING CENTRE. Standard Fortran programming manual. NCC, 1970

NATIONAL COMPUTING CENTRE, Standard Fortran programming manual. 2nd edition.

NCC, 1972.

PRAGER, W Introduction to basic Fortran programming and numerical methods.

Blaisdell, 1965.

RADFORD, A.S.: Teach yourself computer programming/ Fortran. English UP, 1975.

ROHL, J.S.: Programming in Fortran: a course of television lectures.

Manchester UP, 1973.

SASS, C.J.: Fortran IV programming and applications. Holden Day, 1973.

SCHICK, W. & MERZ, C.J.: Fortran for engineering. McGraw H111, 1972.

SLATER, L.J.: Forst[sic] steps in basic Fortran. Chapman & Hall, 1971.

STUART, F.: Fortran programming. Wiley, 1969.

SWEENEY, R.B.: The use of computers in accounting (Fortran). Prentice Hall, 1971.

VINCE, J.A,: Fortran for students. Enfield College Computer Centre, 1968.

WATTERS, J.: Fortran programming: a complete course in writing Fortran programs.

Heineman, 1963.

WILF, H.S.: Programming for a digital computer in the Fortran language. Addison

Wesley, 1969.

VELDMAN, D.J.: Fortran programming for the behavioral sciences. Holt, Rinehart,

& Winston, 1967.

MOON, B.: Computer programming for science and engineering. Butterworths, 1966.

DORN, W.S.: Mathematics and computing: with Fortran programming. Wiley, 1967.

JAMES, M.L., SMITH, G.M. & WOLFORD, J.C.: Appljed numerical methods for digital

computation with Fortran, International Textbook Co., 1967.

MCCORMICK, J.M. & SALVADQRI, M,G.: Numerical methods in Fortran. Prentice Ha1], 1964

MCCRACKEN, D.D. & DORN, W.S.: Numerical methods and Fortran programming with

applications in engineering and science. Wiley, 1964.

BORK, A.M,: Fortran for physics. Addison Wesley, l967.

GENNARO, J.J.: Computer methods in solid mechanics.  Macmillan, l965.

ANTON, M.B. & BOUTELL, W.S.: Fortran and business data processing. McGraw Hill, 1968

CLEVELAND, S.T.: Fortran 2-D on the IBM 1620: introduction to digital computer

programming. Dickenson, 1968.

BLATT, J.M.: Introduction to Fortran IV: using the WATFOR compiler. Goodyear Pub

Co., 1968.

COUGER, J.D. & SHANNON, L.E.: Fortran IV: a programmed instruction approach.

Irwin-Dorsey, 1968.

[The original typescript changed font at this point.]

CRESS, P.:, DIRKSEN, P. & GRAHAM, J.W.: Fortran IV with WATFOR. Prentice Hall, 1968

CRESS, P., DIRKSEN,P. & GRAHAM, J.L.: Fortran IV with WATFOR and WATFIV.

Revised edition. Prentice Hall, 1970.

DOWDING, B.J. & ROGERS, C.B.: Basic Fortran IV. Blackie, 1970.

EDWARDS, P.: Flowcharting and Fortran IV. McGraw Hill, 1973.

HOLDEN, H.L.: Introduction to Fortran IV. Macmillan, 1970.

JAMISON, R.V.: Fortran IV programming: based on the IBM System lI30. McGraw Hill,

1970.

LEDLEY, R.S.: Fortran IV programming. McGraw Hill, 1966.

MCCAMERON, F.A.: Fortran IV. Irwin, 1970.

AMNNING, W.A. & GARNERO, R.S.: A Fortran IV problem solver. McGraw Hill, 1970.

MOURSUND, D.G.: Problem analysis and solution using Fortran IV. Wadsworth

Publishing Co., 1970.

NOLAN, R.L.: Fortran IV computing and applications. Addison Wesley, 1971.

NYDEGGER, A.C.: An introduction to computer programming with an emphasis

Fortran IV. Addison Wesley, 1968.

ORGANICK, E.l. & MEISSNER, L.P.: Fortran IV: standard Fortran Watfor-Watfiv.

2nd edition. Addison Wesley, 1975.

ORGANICK, E.I.: A Fortran IV primer. Addison Wesley, 1966.

POLLACK, S.V.: A guide to Fortran IV. Columbia UP, 1965.

SCHONBECK, R.G.: Fortran IV for multi-programming systems: with emphasis on the

GB-600 series computer. Addison Wesley, 1968.

SILVER, G.A.: Simplified Fortran IV programming. Harcourt Brace Jovanovich, 1971.

TOMASSO, C.: A practical course in Fortran IV. Foulsham, l97l.

[In the original typescript Appendix D was squeezed onto the bottom of the last

page of Appendix A in order to save paper (and possibly postage costs).  In

this electronic version it has been moved to follow Appendix C.]

APPENDIX B

Preprocessor Working Party

Now that a growing number of programmers appreciate that Fortran source

code can be handled as a text, to be generated, edited and otherwise

manipulated, the next few years should see a great deal more software of this

kind produced as an aid to Fortran program development, transfer and maintenance.

It is important to facilitate and learn from these experiments by better

communication, by devising standards for interfaces, and by organizing or

participating in user trials.

Four tasks, which the Preprocessor working Party will undertake, are

as follows.

1.    Study of existing preprocessors in order to find a convenient

vocabulary and classification for discussing them.

2.    Maintenance of a list of available preprocessors.

3.    Be concerned with the opportunities for beneficial new standards or

recommended conventions.

4.    Compile a bibliography of papers helpful to implementors and users of

preprocessors.

Provisional List of Processes that manipulate Fortran Source Text

Appendix C

Summary of Talk by Dr. D. Clifford on DEC Fortran Systems

The talk was mainly about the Fortran Systems available of the PDP8 and PDP11

series computers with 16K words or more. The PDP8 has compilers for Fortran 2

and Fortran 4 with extensions for PHA Sampling routines, Plotter/printer and

Laboratory Experimental Systems. The PDP11 has compilers for Fortran 4 and

Fortran 4+. There exists upwards compatibility from PDP8 Fortran 4 to PDP11

Fortran 4 and 4+. If the ANSI standard is observed, programs are portable from

PDP11 to PDP8. On the PDP11 Fortran 4 can run under the RT11, RSXllM and RSTS E

operating systems, Fortran 4+ operates under the RSX1lD operating system. A

major implementation difference between 4 and 4+ is that Fortran 4 produces

threaded code and Fortran 4+ produces in-line code. Both compilers conform to

the ANSI 1966 standard, and include a few extensions. Fortran 4+ performs quite

sophisticated optimisations including: compile-time evaluation of, constant

expressions, constant conversion and constant subscript expressions, argument

list merging (when two subprograms reference the same arguments, a single copy

of the argument list is generated), logical and arithmetic branch instructions

are optimised often to a single machine instruction, unreachable code is

eliminated, common subexpressions are recognised and replaced, invariant

computations are removed from DO loops, register allocation minimises load

and store instructions and frequently used variables in DO loops are held in

registers. Many of the intrinsic functions (eg ABS) are coded in-line. A

floating point processor is available and the PDP11/45 has hardware

instructions to perform loading/storing jointly with integer/real conversion

and can use core locations in the same way other processors use their

accumulators, again eliminating the need for repetitious load/store for simple

statements like: J=J+K, J=K, I=I*2, K=K-l, J=J+l, I=0, X=0.0 and L=-L which
are converted to a single machine instruction.

Extensions to the language include ENCODE/DECODE, TYPE/ACCEPT and random access

input/output, mixed mode arithmetic expressions and no limit on the number of

continuation lines.

Supporting facilities include a comprehensive scientific subroutine library.

with the RT-11 (real-time) operating system packages for graphics (VT11),

printer plotter (LV11), laboratory peripheral system (LPS11), laboratory system

(AR11) and digital I/O (DR11K) are provided. With the RSX-11 operating system

packages for industrial control and DEC laboratory control interfaces, RSX-11

system directives and I.S.A. standard routines for tasking, bit handling and
process control are provided.

An independent comparison by the CCA using the WHETSTONE benchmark included

the DEC PDP11 and a summary of performance in Kilo-Whetstone-instructions/
second is given below.

CCA BENCHMARKS

WHET1A               WHET2A

SINGLE-PRECISION     DOUBLE-PRECISION

APPENDIX D

         = F O R - W O R D = >

    FORTRAN DEVELOPMENT NEWSLETTER

Volume 2, Number 2         [Pages 11-16]                April 1976

Published by Ad Hoc Committee on Fortran Development, ACM-SIGPLAN

Prepared for the U.S. Energy Research and Development Administration under Contract W-7405-ENG-48

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

X3J3 'FAVORABLY DISPOSED' TOWARD IF-THEN-ELSE IN CURRENT REVISION

In an unofficial action, the Fortran Standards committee (X3J3) expressed its inclination in fa-

or of a proposal to incorporate an IF-THEN-ELSE structure into the current revision of Standard For-

tran.  This was expressed by an informal vote during the X3J3 meeting at NBS Gaithershurg Md (March

31  - April 2), and was no doubt influenced by public response at the Fortran Forum earlier in the

week. A number of members of the committee felt concern because of the "lateness of the hour," but

the prevailing sentiment was that the importance of this feature warranted giving some extra consid-

oration to its adoption. The proposal was presented by Walt Brainerd and is of course subject

to further revision prior to its possible formal adoption.  Public

comment on this proposal is solicited, and should be directed to Lloyd W. Campbell, BRL-CSD Bldg 328,

Aberdeen Proving Ground, MD 21005.