Recommendations
On Splitting the Laboratory
Consultant’s Report (Extract)
Letter: H. Whitfield to V.C.
Paper by H. Whitfield and E.D. Barraclough
[When this Working Party, which was chaired by Professor Alex Crombie, first submitted its Report to Senate, its recommendation that there be a formal separation of the computing service from the academic department of computing, a recommendation that ran counter to the advice contained in the Report of the Committee’s Consultant, was rejected by Senate. Two years later Senate reconsidered this issue, and accepted the recommendation.]
UNIVERSITY OF NEWCASTLE UPON TYNE
WORKING PARTY ON ACADEMIC SERVICES
REPORT TO SENATE ON THE COMPUTING SERVICE
19 April 1989
SUMMARY OF PRINCIPAL RECOMMENDATIONS
(a) That a centralised computing service should be retained. (8(a)).
(b) That special protection from the current financial cutbacks should not be given to the computing service, but that it should receive “average” treatment. (8(c)).
(c) That a new Computing Service Committee should examine future computing needs within Faculties in consultation with the computing service within a period of six months. Underlying this investigation would be the express intention of re-orienting staffing support from the central service to provide more localised support in conjunction with identified complementary resources from Faculties. (8(d) (iv)).
(d) That the allocation of costs is a major policy area which requires careful observation and consideration by the Computing Service Committee. A greater proportion of computing costs should in due course be recovered from users. (8(e)).
(e) That close attention should be given to a greater recovery of computing service costs from research grants carrying overheads. (8(e)).
(f) That consideration be given to the future of computer user groups. (8(f)) .
(g) That the computing service should be formally separated from an academic department of computing. (8(g))
(h) That the formal NUMAC arrangement should be reviewed. (8(h)).
(j) That the University should pursue a more co-ordinated strategy on Information Technology. (8(j)).
(k) That the University should clarify its policy on the funding of the Ethernet installation. (8(k)).
(l) That a new Computing Service Committee be established. (8(m)).
Extract from pp.13-14
(g) (i) Considerable attention has been given to the advantages and disadvantages of the constitutional relationship between the computing service and the academic department of computing. This University is one of three which maintain a structure based on a single department. The basic objectives of academic computing are to advance teaching and research in the discipline, whilst the service is required to perform a service for the University as a whole. The working Party’s attention has been drawn to the collaborative and “sparking-off” elements between the two sections. Whilst welcoming this situation, the Working Party believes that such relationships can be maintained under either structure. It believes that the efficient operation of the Computing Laboratory has depended more on the excellent working relationship which exists between the Director and the Executive Director than on the management structure which is ambiguous and insufficiently well-defined.
(ii) The Working Party therefore recommends a formal separation of the computing service from the academic department of computing. The benefits it perceives are:
(1) A reduction in the current heavy workload carried by the Director of the Computing Laboratory.
(2) A clearer identification of the status and responsibilities of the Director of the Computing Service.
(3) A clearer indication that the Director’s primary responsibility rests with the development of a service responsive to the expectations of its users across the University and which is more “consumer-driven” than at present.
(4) Clear line management of the Computing Service through a Director reporting to a Computing Service Committee and a Spending Authority.
(5) Clear line management of the academic department of computing through a Head of Department in the Faculty of Science, in accordance with the University’s normal arrangements.
(6) Clearer line management within the Computing service below the level of Director, with existing reporting relationships being the subject of future review.
(7) Improved financial accounting and apportionment of costs as between the two areas of activity.
(8) Clear governing structures above Director level, particularly in regard to the determination of general computing service policy within the University at large.
[EXTRACT]
CONSULTANT’S REPORT ON THE COMPUTING SERVICE
D. F. Hartley
February 1989
27 It is now unusual for Computing Science and a Computing Service to be organised in a single department, although three U.K. universities: Cambridge, Kent and Newcastle have maintained such a structure.1 But all three are recognised to be amongst the leaders in both academic and service affairs. It may not be easy to determine the extent to which past and present successes are due to ad hominem factors, but I have no doubt that, given the correct structures, disadvantages can be avoided and major gains made. The details of the arrangements between Cambridge and Newcastle are very similar, and there is only one significant difference which I comment on later.
28 Reasons for fearing that such an arrangements may lead to disadvantages are based on the fundamental difference between an academic department, whose objectives are to advance the research and teaching of a specific discipline, and a service organisation whose duties are to meet the needs of the university a whole. The objectives are manifestly different, and it is therefore important that there be no conflict between the two. I believe that this requires clear definitions in the areas of (i) senior management, (ii) the duties of individual staff, (ili) financial accounting and (iv) the governing structure above the department, particularly that concerned with the determination of Computing Service policy.
29 Arguments in favour are invariably founded on a history of past successes, and include the benefits of intellectual influences flowing (in both directions) between science and service. For example, we have found in Cambridge that, although Computer Science now operates a substantial amount of their own equipment, they continue to make significant use of central service facilities, thus avoiding unnecessary duplication and providing a useful user-influence on the Computing Service. Both types of benefit work in the opposite direction: the Computing Service’s more pragmatic approach brings a steadying influence to Computer Science, while the relatively advanced facilities and resources of the latter can used, through the Computing Service, to the benefit of the university generally.
30 I was informed of specific instances where there is real mutual benefit from the present arrangements at Newcastle. For example, the Computing Service currently must widen its experience away from the dominance of the MTS operating system, and justifiably needs to explore the potentialities of the Unix operating system. Computing Science has substantial Unix expertise, and their support staff are helping the Computing Service without detriment to either side; but for this the Computing Service would have a strong case for additional staff. In addition economies accrue from the sharing of common administrative and technical support staff. I have no doubt in my mind that a decision by the University, for whatever reason, to divide the Computing Laboratory into two separate departments would lead to a permanent and significant increase in costs.
31. I saw no evidence of real disadvantages arising in the present relationship: the position of Miss Barraclough as Executive Director gives a direct line of responsibility under her for Computing Service affairs, and the position of each member of staff seems adequately clear. To the extent I have investigated the point, the financial accounting also makes a proper distinction between academic and service expenditure. Academic policy is well catered for under the Faculty of Science, while the Computing Laboratory Committee essentially concentrates on Computing Service affairs.
32 Much of the Newcastle structure is reflected in the arrangements in Cambridge, which also work well and are seen to be politically acceptable in that University. The main difference is in the senior management: the Head of Department at Cambridge is an elected senior academic who has responsibility for the Computing Service in an administrative sense only, and not for the nature and quality of its activities. The Director of the Service is wholly responsible (to the equivalent of the Computing Laboratory Committee) for Computing Service management and performance; his responsibility to the Head is in a domestic sense only. This is to be contrasted with Newcastle, where the Head of Department is responsible for both sides and there is a ‘one over one’ relationship between him and the Executive Director. While there is no evidence whatsoever of this currently causing any problems, it must in theory be problematical, and could cause difficulties in future whenever either position have to be refilled.
Footnote 1
I understand that Lancaster has recently re-organised its computing along similar lines, but I have no knowledge of the details.