| QUOTE |
| [url]www.newcastle.research.ec.org/.../ news/issue5.htm[/url] 6 Spotlight on... Brian Randell Brian has announced his retirement with effect from 30th September 2001. He came to Newcastle University in 1969 as Professor of Computing Science and in 1971 set up the project that initiated research into the possibility of software fault tolerance, and introduced the "recovery block" concept. Subsequent major developments included the Newcastle Connection, and the prototype Distributed Secure System. He has been Principal Investigator on a succession of research projects in reliability and security funded by the Science Research Council (now Engineering and Physical Sciences Research Council), the Ministry of Defence, the European Strategic Programme of Research in Information Technology (ESPRIT), and now the European Information Society Technologies (IST) Programme. We all hope that he will continue to be involved in CaberNet and the IST research projects DSOS and MAFTIA for a long time to come. The following is provided by Dr C.R. Snow, the new CaberNet Director. Brian Randell's technical work in Dependability of Computer Systems is well known, and well documented in his many publications. What is perhaps less well known is the role he had in the facilitating of basic research in computer science through funding from the European Commission. In the mid- to late- 1980s, there was considerable uncertainty in the UK computing community regarding the way in which research would be funded after the end of the Alvey project, and Brian had some preliminary conversations with the then Head of ESPRIT. Some time later Brian met George Metakidis, who worked for the Commission, and who, probably largely as a result of Brian's persistence, had been charged with looking into the possibility of basic computer science research being funded by ESPRIT. Thus was ESPRIT BR conceived. (To avoid confusion, we have to emphasise that BR stands in this case for Basic Research, and not Brian Randell!) The (political) collaboration between Brian and George Metakidis continued, and at some later point they hit upon the notion of forming a loose association of research groups in a particular area of expertise. On being asked to identify an area, Brian, predictably, said "Dependability and Distributed Systems". Out of this the whole concept of the "Network of Excellence" emerged. George had managed to lay hands on sufficient funding to set up three pilot networks, and Brian, working with the many contacts he had in the field of Distributed Systems and Dependability, put together a proposal for a Network of Excellence in Distributed Computer Systems Architectures. The proposal was created, refined and submitted in the incredibly short time of two weeks. Such is the power of e-mail! The Network of Excellence concept was designed to encourage fora in which related research could be discussed before the formal creation of a project. A project, by its very nature, has a particular focus, and a set of research goals, not to mention deliverables. The idea of the NoE was to allow groups to discover what work was going on in different but related areas, in the hope that new collaborations would occur, and new projects created. Three pilot networks were created, ELSNet - the network for language and speech, CompulogNet - computational logic, and DCSA. The name DCSA did not last long. Soon the rather more "catchy" name of CaberNet emerged. As with all good acronyms, the acronym came first, and the interpretation followed. The "Net" part was easy, and so was the "CA", standing for "Computer Architectures". The "BER" required a little more thought, but when it was realised that much of CaberNet's work, and particularly its proposed infrastructure, could be usefully employed by other Networks of Excellence, it did not seem too presumptive for this to mean "Basic European Research" - hence CaberNet, the Network of Excellence in Computer Architectures for Basic European Research. Initially, whilst Brian remained heavily involved, he felt that for a number of reasons he did not wish to act as CaberNet's Director, and persuaded Professor Harry Whitfield and myself to become involved, as Director and Deputy Director respectively. The pilot Networks lasted for one year, and CaberNet was subsequently funded for two further three-year periods. Brian continued to provide inspiration in the work of the Network, and sometime during 1992, he and Jean-Claude Laprie were mulling over ways in which the Network could evolve, and be more pro-active in promoting research the the areas which were close to both their hearts. During the course of these conversations, the concept of the Young Turks movement emerged. However, given that one of the members of the Executive Board was Turkish, and that the thought that he might be considered an "Old Turk", it was felt that the "Young Turks" label should not be used, but became known as the "Radicals". The radicals idea was to invite CaberNet nodes to nominate one, or possibly two, of their bright young researchers - recent PhDs for example - to attend the Radicals Workshop. This forum was deliberately designed to allow these young researchers to discuss together, without the inhibiting presence of senior members of their respective research groups. The only "senior" member of CaberNet who was permitted to attend was the project Administrator, who provided administrative advice to the attendees regarding the bureaucratic aspects of European research. The expectation was that these workshops would stimulate new collaborations and lead to new project proposals. In 1994, for a number of reasons, Harry Whitfield decided that he wished to resign as Director of CaberNet, and Brian was asked if he would now formalise his involvement in the Network by becoming its new Director. Brian agreed, and at an Executive Board meeting held in conjunction with the Network's Plenary Workshop in Dublin, it was agreed unanimously and with acclaim that Brian should take over the Directorship, and Chairmanship of the Executive Board. It was clear at a reasonably early stage that much of CaberNet's work, particularly in respect of the infrastructure required to allow the nodes to collaborate, could also be of great value to the other Networks of Excellence. It was also realised that there could be some value in Networks co-operating with each other in, for example, their dealings with the European Commission. Brian was instrumental in arranging a number of meetings at which all of the Networks of Excellence were represented, and at which Networks were able to learn about each other's requirements and procedures. This proved to be highly beneficial certainly in respect of CaberNet's dealings with Brussels, and one hopes that the other Networks felt similarly. CaberNet continued its work under Brian's leadership, but the project came to an end (i.e. ran out of funding) at the end of 1998. A non-funded extension was applied for, and granted, for a further six months, but for eighteen months the Network remained in a state of limbo, while Brussels decided what future the Networks of Excellence should have. Eventually, decisions were made, a new proposal written and approved, and the latest phase of CaberNet began on January 1st this year. At the first Executive Board meeting of the new project, held in Newcastle in early February, Brian announced that, in view of his impending "retirement" from the University, he wished to stand down as CaberNet's Director. The Board was not surprised by this news, but with great reluctance accepted it. This has necessarily been a somewhat "potted" version of the influence Brian has had on the politics of computer science research, and has concentrated on the way in which that effort relates to CaberNet. There is no doubt that there is much more to say about Brian's impact on the international computer science community, but that would probably require a fully fledged biography. Earlier, I placed the word "retirement" in inverted commas. Brian himself has described it as a "virtual retirement". He will still keep his office (for the time being, at least) in the Computing Science Department at Newcastle University, he has agreed to be involved with certain aspects of the life of the Department, and there is little doubt that he will continue to contribute technically to research in the Dependability in Computing Systems, which has been such a large part of his professional life. I should also be surprised if his contribution to CaberNet did not continue, albeit no longer in so prominent a manner. Certainly I, who feels privileged to have inherited the mantle of CaberNet from Brian, will have little hesitation in consulting him whenever I see the need. And knowing Brian, his advice will always be freely available and well worth having. Brian, we thank you for all of this work, and especially for your contribution to the establishment and smooth running of CaberNet over the past ten years. We shall endeavour to ensure that the Network continues to thrive and to provide the opportunities for high quality research in Dependability, just as it has done in the past under your guidance. |
| QUOTE |
| SRL is a somewhat unorthodox place, populated by strange and dubious denizens. |
