The minutes of the meetings held on 3 June and 24 August were approved.
There were no matters arising from the minutes.
The paper gave the standard tables of operating statistics; Dr Howlett gave a verbal statement on the various activities in the Laboratory, which will be issued as an Addendum to ACC 70/7.
The Committee noted the very stable state of the machine performance and the work load; Dr Howlett said that the level of the load was not a true measure of demand, because when a long queue had built up (as happened from time to time) users became discouraged by the slow turn-round and stopped sending work. Also they knew they could not get very large amounts of time and therefore did not submit the kind of project which would need, say, 2-3 hour runs on a machine in the CDC 7600 class. Apart from the delay to long jobs when a queue had built up the pattern of turn-round remained fairly stable, with about 60% sent out within 24 hours of arrival, 80% within 48 hours and 99% within a week.
Dr Howlett reported the election of Dr Churchhouse to the newly-created Chair of Computing Mathematics at Cardiff, which he would take up in February 1971; and the recent tragic death of Dr Nigel Campbell, a young and able numerical analyst in the Laboratory, in an inexplicable accident.
The Committee felt that it did not have enough information, particularly on pressure for space in the existing building, to decide whether or not this proposed building was necessary. It recognised that in authorising the 1906A it had by implication committed the Laboratory to an increase in staff - although many of these additional people would be housed in the new machine block - but felt that what was being asked for was expensive in relation to the number of people to be housed. Dr Howlett was asked to go into the details of the proposal with SRC officials and to bring the proposal back, revised if necessary, at the next meeting.
The Committee accepted all the arguments given in the paper, giving particular weight to the need to protect the file store. They approved the purchase of the two 160 Kch/sec 9-track decks at £35,000 and the two 60 Kch/sec 7-track decks at £26,500; but asked Dr Howlett, before taking the, matter further, to look into the possibility of using decks made by some specialist manufacturer (eg Memorex) instead of ICL.
Dr Howlett emphasized the point made in the paper, that the existing cameras could only with great difficulty be maintained in precise enough adjustment to produce films of the quality needed. The Committee recognised the value of cine films in showing the behaviour of complex dynamical systems, but did not feel convinced that the demand in the Laboratory was great enough, nor the drawbacks of the present camera severe enough, to justify the high cost. Dr Dunworth said that he would like to make some enquiries into the possibilities for this technique and raised the question of whether there was any likely commercial interest. Dr Howlett was asked to provide more information at the next meeting.
The Committee noted that these would be considered in detail by the Finance Division, along with all the estimates put to the Council. Whilst recognising that the aim to replace Atlas with a new major machine in 1973 implied a very short time scale, they agreed that the paper should go to the next stage in the financial programme.
The Chairman opened the discussion on this paper by saying that, whilst he was in sympathy with one of the key points, namely that applications for large amounts of machine time should go through the appropriate SRC subject committees like all other grant applications, he would not wish this to have the effect of concentrating the Laboratory's resources too much on the subjects selected by the Council for special support. Dr Howlett confirmed that this was not his intention, which was that all applications should be judged solely on their scientific merits and given the support which they rated. Mr Jolliffe supported this, saying that whilst SRC were indeed encouraging development in certain selected fields, as described in a policy statement recently circulated to all university science departments, they were putting only about 20% of their grant funds into these favoured fields. He said that the type of computational support proposed in the paper was a proper function of the SRC, whereas the provision of a general service was the function of the Computer Board, not the SRC; and felt that the proposals were more likely to receive approval from DES and the Treasury than a continuation of the general service.
Mr Rutterford, on behalf of the Computer Board, said that whilst the Board had been able to have only a brief discussion of the proposals, they welcomed them. The Board had always valued the services provided by Chilton very highly and wished to see these continued; they welcomed particularly the paper's proposals for software development and could see possibilities for valuable contributions to the 1906A programme. The Board were conscious of a change in the Laboratory's position as a result of the installation of large machines in several universities and were looking to the SRC to define its (the laboratory's) role.
Miss Bowell said that the (SRC) Computing Science Committee had set up a Panel, which would shortly be having its first meeting, to look into the needs for applications software and ways of getting this written. They would be glad to have the Laboratory's collaboration and support.
In the general discussion the following points were made:
The Committee gave their general approval to the proposals of the paper and the Chairman authorised Dr Howlett to submit them to the Science Board at its meeting of 9 December 1970. Mr Jolliffe said that as the services provided by the Laboratory were of interest to the other Boards, particularly the Engineering and ASR Boards, the views of these would be sought.
