Contact us Heritage collections Image license terms
HOME ACL Associates Technology Literature Applications Society Software revisited
Further reading □ Overview07/07/6120/10/6116/02/6206/07/6227/11/6229/05/6313/11/6306/05/6408/10/64
ACD C&A INF CCD CISD Archives Contact us Heritage archives Image license terms

Search

   
ACLLiteratureCommittee MinutesNIRNS :: NIRNS ACC Minutes
ACLLiteratureCommittee MinutesNIRNS :: NIRNS ACC Minutes
ACL ACD C&A INF CCD CISD Archives
Further reading

Overview
07/07/61
20/10/61
16/02/62
06/07/62
27/11/62
29/05/63
13/11/63
06/05/64
08/10/64

Minutes of the meeting held on 27/ 11/1962 at 11, Charles II St, London

1 CONGRATULATIONS

Through the Chairman, the Committee warmly congratulated Dr Kendrew on his Nobel prize.

2 MINUTES OF THE LAST MEETING

The Committee approved the minutes of the meeting on 6th July, 1962 subject to an amendment in the last sentence of Minute 7.1, where it was agreed that the words as a supplement should be replaced by the words in addition.

3 REPORT BY DR HOWLETT

The main points of Dr Howlett's report were as set out in paper NI/ACC/62/8. Further detail and discussion of various points were as follows:

3.1 The Computer: It was noted in particular that Ferranti's reported the main store to be behind schedule and likely to be the main holding factor on commissioning the machine. This wording was taken to indicate a possibility of delay in the programme, and Dr Howlett was asked to keep strong pressure on the contractors. The latter were aware that any delay on this machine would be a severe blow to their prestige as well as delaying the start of scientific work.

3.2 Direct Data Input: Dr Howlett said that he had arranged a one-day symposium to collect information from people who would like a fast data input facility to be added to the Atlas, so that it could be used for computation on-line with certain experiments. Ferrant's had said informally that they could provide this in about two years at a cost of the order of £50,000. In discussion, the Committee recognised that this suggestion involved the general question of use and priorities of the Atlas. It is not axiomatic that direct linkage with large-scale experimental equipment is appropriate for a machine which is intended to provide a general and central computer service; but if the suggestion would greatly increase the capability of the machine it should obviously be considered. The Chairman drew attention to the activities of various international bodies, in particular the ENEA, in connection with centralised data processing services and said that in any plans of ours we should take note of theirs. Dr Howlett was asked to prepare a technical paper for the next meeting (a) outlining the possibilities, time scale and cost (b) reviewing the advisability of doing such work directly on a central computer, in comparison with the use of smaller, intermediate computers and (c) taking account of any other relevant national or international plans for data processing.

3.3 Programming Languages: It was again pointed out that the AEA, CERN and the Rutherford Laboratory have adopted Fortran while certain other laboratories use Algol and other languages. the Committee reaffirmed their policy that the NIRNS Atlas should be able to accept programmes in any of the major languages adopted by its users. This did not however mean that the Atlas Laboratory programming group would have to be expert in all these and offer a programming service in each.

3.4 Staff: In connection with the report on staff, Dr Howlett said that two consultants had been appointed: Professor M G Kendall and Mr Christopher Strachey.

The Personnel Committee had approved the appointment of the Head of the Programming Group at Band level if necessary (but had asked that the possibility of filling it satisfactorily at PSO level should be tried). An advertisement giving the two alternative salary ranges was about to appear.

3.5 The Edinburgh - Manchester Data Link: The Committee asked Dr Howlett to prepare for a future meeting a technical report on data links from user centres to a large computer, based on the large scale experiments carried out in the AEA and also the experimental Edinburgh - Manchester link to which the Institute had contributed. One point made in discussion was that the latter experiment was useful in having a university at the user end, so that some of the problems of university users in this situation would be explored.

3.6 Joint NIRNS - College Fellowships: The Committee welcomed the idea of joint fellowships partly in the Institute and partly in an Oxford College (for example) but agreed that this was a matter for the Institute to deal with in general rather than in particular relationship to the Atlas Laboratory. Dr Pickavance said that discussions were going on rather cautiously as there were many pitfalls to be avoided.

4 PROGRESS WITH THE MANCHESTER ATLAS

The Chairman said that Professor Kilburn had written to him saying that the Manchester Atlas project was in a very satisfactory state and giving a detailed summary of the progress and performance. The Chairman arranged for copies of the latter to be sent to all members of the Committee.

5 PROPOSED ADVISORY COMMITTEE - NI/ACC/62/9

After a full discussion, it was agreed that a committee broadly as described by Dr Howlett in his paper was required, but that it should be called the User Committee and should be slightly modified as described in paragraph 5.1 below. It was also agreed that there was an urgent need for policy planning (probably requiring a Policy Committee or thinking committee) on the lines described in paragraph 5.2 below.

5.1 The USERS Committee: It was agreed that in addition to advising the Director of the laboratory on questions of use of the computers, as detailed in the paper, the Committee would also advise and support him when necessary on the question whether particular computations could be done on a smaller machine rather than on the Atlas. (It was also agreed that this necessity would arise less frequently if requests for Atlas time were routed through the head of the university computing group in the university concerned.) Thirdly, the Users Committee would be a forum for any appeals concerning service and arrangements, but not appeals concerning priorities.

The composition of the Committee as proposed in paper 62/9 was approved, subject to an increase in the number of university representatives from one to either two or three. In choosing these university representatives, the advice of the Inter-Universities Computing Committee would be sought. As far as possible, the membership of the committee should be chosen to include representatives of the main user sciences, as well as of the main types of user organisations.

Dr Howlett was asked to spell out the slightly revised terms of reference of the Committee in a new paper.

5.2 Policy and Priorities: It was agreed that a policy should now be worked out concerning the broad allocation of priorities, but that information was first needed about the big computations proposed so that they could be taken into account for planning purposes. Dr Howlett was asked to write to all the users saying that we were trying to formulate our long-term plans and were therefore enquiring about large items of computation proposed. Dr Howlett would invite a statement about any large programmes which might be coming to the NIRNS Atlas together with an indication of priority where one user had more than one large item. Dr Howlett was asked to discuss the replies in 2 or 3 months time with the Chairman who might ask a few members of the Committee to join in. It was expected that it would be necessary to ask many users for further details before a reasonably balanced view of the demands could be obtained.

The Chairman said that if possible he would like this exercise to result in the preparation of a review of the first major requirements to be presented to the Atlas Computer Committee at their May, 1963 meeting. They could then could consider what means to adopt in order to make the broad decision on priorities, and to deal with questions such as computer languages, and whether to provide for on-line computing (Minute 3.2 above). It seemed probable that a Priorities Sub-Committee would be set up meeting more frequently than the Atlas Computer Committee, and having power to co-opt experts in particular subjects whenever necessary. A decision on the matter was however deferred pending the assembly of the information referred to above.

The large number of relatively small computations presented a different problem and this would probably be dealt with by allocating a certain proportion of time to small computations, and leaving it to Dr Howlett to allocate this, reporting periodically to the Priorities Committee for any adjustments that seemed to be required, in the light of the experience gained.

6 UNIVERSITY USE OF AEA COMPUTERS - 62/10

The Committee took note of the difficulties described in paper NI/ACC/62/10, which had resulted in a very great reduction in the scale of the proposal for university use of AEA computers with NIRNS financial support. Dr Howlett said that he would welcome even this reduced scheme as a start in providing a service to Universities. in discussion, however, doubts were raised as to whether even the small amount of time indicated would be available. The Chairman said that a definite answer to this must be given by the AEA by the end of the year.

7 PROPOSED CHANGE AND ADDITION TO THE COMPUTER - 62/11

The Committee considered paper NI/ACC/62/11 and after discussion they most strongly recommended the proposed replacement of the two fast printers by the newly-available Anelex printers and the proposed addition of an engineers Console, at a total additional cost of £27,000.

Although no further additions were yet being submitted for consideration, the Committee had little doubt that this would happen from time to time. two examples were mentioned:

  1. a microfilm output recording either figures or curves. In the latter case the curves could usefully be displayed as movie pictures in some applications, e.g. in the calculation of the propagation of shock-waves.
  2. an 8-track paper tape input. (The standard Atlas tape-reader accepts only 5, 6 or 7-track tape, but many other machines use 8-track tape, including the KDF9's to be installed in several universities.)

These two items were only given as possible examples and it was pointed out that these two would need careful assessment because they would only affect certain users, whereas the proposals just recommended affected all uses of the computer. However, the Committee agreed that in the rapidly developing large computer field, additional equipment would certainly be needed, and they considered that this should be made clear to the Treasury, in order that the Treasury should appreciate the position

⇑ Top of page
© Chilton Computing and UKRI Science and Technology Facilities Council webmaster@chilton-computing.org.uk
Our thanks to UKRI Science and Technology Facilities Council for hosting this site