As reported last month (CI No 1,475), large UK information technology users have got together and formed the Object Interest Group to explore the possibilities offered by object-oriented techniques. The man behind the user initiative is Norman Plant, who is manager of object-oriented design at British Airways Plc. He first became interested in object-oriented techniques a year ago and swiftly set up a task force in British Airways to do pilot (as in test) work in the area. It is fair to say that Plant is an enthusiastic convert to object-orientation, and realised that the new technology’s biggest problem in terms of being widely accepted was cultural rather than technological. He felt that more vendors in the computer industry should be made aware of the technology’s potential, and that it should be popularised among data processing managers that don’t like to take risks with new, relatively untried products.

Credibility

Although Plant felt he was making some headway tackling these issues internally at British Airways, he felt that this wasn’t enough – that what was required was a wider, co-ordinated industry view. To this end he approached data processing managers in 16 companies, including well-known names such as British Telecom, Ford and ICI, to get them to put proposals to their companies enabling them to contribute 40 man days each to the project. The Object Interest Group held its inaugural meeting in January and decided that it should try to establish a shared assessment of the credibility of object-oriented technology for use in the 1990s. In order to do this it had to involve major vendors; for, although Plant knew that the majority of large vendors were already exploring object-oriented techniques, he felt these companies had pockets of interest in the technology, but no clear focus or direction. Therefore, the Object Interest Group believed that its presence would cause these pockets of interest to coalesce and encourage the evolution of vendor policy. Therefore, 18 vendors were approached this summer to take part in the Group’s project. Vendors approached include Oracle, Sun, Data General, Unisys, Microsoft, Logica, ICL, Olivetti, DEC, Hewlett-Packard, Ontos, IBM and AT&T. British Airways had the responsibility for sounding out IBM, and Plant says that IBM UK’s managing director Tony Cleaver was very helpful. For example, Cleaver appointed a manager to liaise with the Group, IBM has arranged a seminar for the Group defining IBM’s position, and IBM is also setting up visits to the US for the Group’s representatives so they can meet the people responsible for Systems Application Architecture. –

By Katy Ring

However, Plant emphasised that the Group is not affiliated to any companies or movements within the computer industry and intends to retain its objectivity and independence. It is only interested in a shared understanding of object-oriented techniques. Techniques that Plant believes could take software from the cottage industry it is at present into the realms of precise engineering. However, Plant recognises that he is an enthusiast, and so has been careful also to enlist the support of sceptics for the Group. The Group is addressing 16 topics with a standard set of questions and has divided itself into four teams: the first is tackling analysis and design, including software engineering support, the second is exploring languages and hardware systems, the third is investigating database technology and standards groups, while the fourth is looking at project management and training. When the project comes to an end at Christmas, Plant hopes to have achieved a report summarising the Group’s findings, which will be available for all the members of the Group. He is unsure whether it will be made public, since he thinks the Department of Trade & Industry, which is a member member, may want to make the findings widely known, but he knows that most members will keep them for internal use only. At present all the information that the four research teams collect is being entered into a repository under topic

headings. As there will undoubtedly be conflicting views under any one topic heading, team leaders must reconcile these views so that a coherent report can be produced. The Group will also present participating vendors with a statement of direction and demand. But, of most importance, as far as Plant is concerned, is the knowledge and experience which participating individuals will acquire. For, as Plant says, it is individuals and not reports that change companies.

Neutrality The whole project needs to be completed quickly, since the users belonging to the Group tend to belong to rival companies and so could not work together for very long. Once the report is complete Plant believes it may well instigate a big change in the way systems are designed, built and maintained. But companies will then be free to exploit object-oriented techniques the best way they can. The Group is in no sense a lobby or a standards body, although talks are taking place between it and the US Object Management Group. The latter would very much like the UK Object Interest Group to become involved with its User Advisory Group, and this is being proposed at the next Object Interest Group meeting. However, such a move could compromise the UK Group’s neutrality, although some kind of relationship between the two bodies makes sense. There is no user group focussed on object-oriented techniques in the US, and yet there is clearly a need for more projects such as that set up by the Object Interest Group. For as Plant says, only a user could really have a go at something like this because only user companies have no vested vendor interest. And Plant is adamant that object-orientation could well prove to be the panacea that data processing departments have been searching for. As he puts it IT departments get a rough ride in terms of company responsiveness – object-orientation represents a way of addressing all these problems.