OUCS Annual Report
1. General Overview
The chief features of the operation of OUCS during the year ending on 31-Jul-99 are those of continuing a wide range of services, broadening that range of services, and increases in activity in almost every area. Against this background of vigorous service, we have continued to review what we do and how we do it, seeking to respond to a variety of pressures, sometimes conflicting.
These pressures include the ever sharpening need for transparency and accountability, the furious pace of technological change, the ever developing needs, expectations and capabilities and indeed numbers of users of our services, and the imperative to respond to changing organisational structures and politics within the University. It is probably wrong to talk of "users" of our services any longer, since almost everyone in the University now makes use of them, whether consciously or transparently.
One of our goals is to make, as far as we can, basic infrastructure services "invisible" to those who use them. To a large extent, this is being achieved, especially for most of the network services - the Backbone, the Janet connection, Email relays, Web servers, and so on. We take some pride in this achievement. However, one unexpected and undesirable consequence is the increasing invisibility of the cost of doing so, of the hard work and expertise of key individuals, and of the extent of the University's vulnerability to these services on which it has, quietly and perhaps unwittingly, become extremely dependent.
It should be noted that the Backbone, for instance, is critically dependent on a few key technical experts. We are indebted to such staff for their high degree of competence, commitment and loyalty, in the face of very competitive employment opportunities outside the University. Our vulnerability is doubled because we operate on far lower staffing levels than typical commercial enterprises with similar degrees of dependence on their IT infrastructure. Users of these services and the University as a whole must therefore recognise that the desired (perhaps even necessary) level of guarantees regarding reliability, response times and performance can never be furnished at current resourcing levels.
We seek to support and encourage and reward these staff members (and a few others in other key areas) as much as we can within the personnel systems operated by the University. We believe that, while at the moment the staffing situation is stable, we are nevertheless extremely vulnerable, and there is little we can do within the current structures to reduce that vulnerability.
It must also be noted that no funding is provided (and no payment to staff is permitted) for any formal support "outside hours". Many of the staff responsible for critical areas undertake voluntary monitoring of the services (by dial-up from home), and report faults that they cannot fix themselves to those responsible, or to management. However, if staff are away from Oxford, or have unavoidable commitments, there may be little that can be done immediately. OUCS, and ultimately the University, relies heavily on the goodwill of all its staff, and is anxious not to dissipate that goodwill by increasing demands that would, in most environments, already be considered to be unreasonable. No solution to this dilemma is apparent.
Although staff numbers at OUCS have remained constant for some time, apart from new externally-funded or self-financing activities, we have coped with the substantial increases in load by a variety of means. Tribute must first be paid to all staff who have typically taken on extra load wherever possible, and all are now operating at exceptional levels. This takes its toll: in some cases, important work has been delayed or has had to be deferred (eg introduction of a Firewall, and systems to make rapid identification of traffic anomalies more automated); in some cases operational work has been undertaken by managers, which limits their ability to do their job properly; in the longer term, these added pressures may make it harder to retain and recruit competent staff. In all cases, there is now insufficient provision for contingencies, reflection, and for an adequate amount of experimentation and directed research into new technologies and services. This is barely noticeable in the short term, but could have very serious consequences in the long term, as the ability of the University to adapt to or exploit emerging technologies is severely hampered.
Increased use of all OUCS facilities and services places burdens not only on those who directly provide them, but also on those who support them, who train people in their use, and who deal with the constant flow of queries concerning their use.
Another way in which we have responded to the increasing loads has been to review our systems and procedures, wherever possible improving their efficiency or introducing cost-effective automation (eg on-line registration systems, and pre-registering all new undergraduates with email addresses). This has required an increased reliance on possession of the University Card for identification and validation. But again, our ability to achieve this has been hampered by operational loads ("it's difficult to remember as you fight off alligators that you set out to drain the swamp!").
We are very conscious that even declining services fail to contribute much to reducing costs or loads, since a little-used service can still be vital to those using it, and "little used" means now what full usage meant a few years ago, given the enormous widening usage of IT which we have experienced over the past 5 years. Nevertheless, we have also responded to changing demands by redeployment and training of staff wherever that has been possible.
Overall, we are convinced that we render an exceptional IT service to the University, and this is often borne out by the many compliments we receive, especially where we are compared favourably to the service provided at other universities, whether in the UK or the USA. Of course, we do receive complaints from time to time (though not actually very often), and we take these seriously in our endeavours to improve our services. But the fact remains that we believe that OUCS provides a service appropriate to the international standing of Oxford University. Oxford has IT services and facilities (not all provided by OUCS, of course) that are the equal of any in the UK, and even of many of the better-endowed US institutions. We have capable and sound IT infrastructures; broadly speaking, we have good support structures in place; and we have ample provision of networked computers - probably roughly 1:1 for staff, and better than 1:5 for students (excluding students' own computers, which must now exceed 1:2).
Although the fact that our cost of trans-Atlantic Janet traffic is about the greatest in the UK brings its own problems, it is a reflection of the highly skilled and well-equipped staff and student body, in IT terms, of this University. We have a very good platform from which to exploit the benefits of IT more effectively. It is clear that there are many vital opportunities for the use of IT which the University is now in a good position to exploit in its research and teaching activities, both operationally and at a more strategic level.
Much of the gain is attributable to sound overall policies on the provision of IT (through the IT Strategy which has been kept under review, with revisions in 1996 and again this year), the use of devolved responsibility coupled with sound infrastructure and sensible central initiatives. One of the more successful initiatives (apart from the network) over the past few years has been the Software Site Licence Scheme, which never ceases to achieve plaudits from staff in every part of the University. The extension of this to encompass the Microsoft Campus Agreement scheme, while a major development, has built on these very successful arrangements. It was undertaken only after a very extensive consultation process, which has helped to achieve acceptance of the outcome.
The long-term question of (perhaps increasing) reliance on Microsoft's products is something that has stirred a considerable level of feeling, not just in this university. Some people believe that an "open source" solution is to be preferred (eg using a product like Star Office on a Linux platform), since it is non-proprietary and seemingly far less expensive - even setting aside ideological questions about multi-national corporations with near-monopolies. It is clear that these issues will continue to rise, but as long as the vast majority of the University wants to make use of proprietary products, then OUCS must do all it can to facilitate ownership and support. For those wishing to investigate alternatives further, the JISC has prepared a Report on the "cost of ownership" of Office software (see http://www.jtap.ac.uk/reports/htm/jtap-029.html), UCISA intends to run a workshop on the same issue, drawing on this Report, and an eminently readable book on the whole question of open-source software has recently been published ("The Cathedral and the Bazaar", by Eric S Raymond, O'Reilly, 1999).
OUCS ended the year with an operational deficit of £138,042 (excluding externally-funded activities), on an overall turnover of £6,037,692. The main grants from the General Board amounted to £3,582,610, with the balance of revenue coming from charges and earned income. This situation is satisfactory, given a few once-only expenses, which were taken from our reserves.
We were all shocked and saddened by the sudden death of Rob Hutchings in a road accident on 7-Aug-98. Rob had been with OUCS since 1971, most recently as an expert in typesetting and text formatting languages, especially Postscript and TeX. We have not effectively been able to replace this expertise. We have extended our condolences to members of his family.
Another departure under much less traumatic circumstances was the retirement of Keith Moulden, our Administrator for the last 9 years. He officially finished on 30-Sep-98, but he continued to help us out on a casual basis during the period in which we were unable to find a suitable replacement in the short term. We have wished him every benefit of his retirement. We were without the full-time services of an Administrator for 7 months, until the arrival of Karen Young on 17-May-99, who transferred from the University Chest.
Sue Brooks' arrival, being on 14-Apr-98, was actually in the last reporting period; but this has been her first full year as Assistant Director (Distributed Computing). She was "thrown in at the deep end", by being asked to draw up plans for the revisions to the Help Desk service, which consumed enormous amounts of her time, energy and creativity. As a consequence, she has barely had time to develop some of the distributed computing aspects of her post. Nevertheless, she has already contributed to OUCS and to the University more broadly in a wide range of areas, and this has been much appreciated.
Overall, all OUCS staff contribute in a myriad of vital ways to the mission of the University of Oxford, and tribute must be paid to their consistently sterling efforts. We are also grateful for the high degree of cooperation which others within the University and Colleges who also provide IT services accord us. Furthermore, although our official status as academic-related and non-academic staff, and our presence on some committees as people "in attendance" rather than as full members, and our status as a "service organisation", may all seem to say otherwise, we do believe that we must and do operate in partnership with academic and other colleagues. Only in partnership can we fully exploit the potential of all staff and of what IT has to offer.