There is widespread agreement within the industry that the key to getting object-oriented technology out from vendor laboratories and into the harsh light of user sites is to win the support of systems integrators. There are stories doing the rounds that some users have requested that mission-critical applications incorporate object-oriented technology, only to find that […]
There is widespread agreement within the industry that the key to getting object-oriented technology out from vendor laboratories and into the harsh light of user sites is to win the support of systems integrators. There are stories doing the rounds that some users have requested that mission-critical applications incorporate object-oriented technology, only to find that the system integrator isn’t prepared to shoulder the responsibility of breaking new technological ground. One consultancy that wants to be identified with object technology is Coopers, Lybrand & Deloitte, which already boasts an Object Technology Services division and, in the US, offers two products in the area: a class browser and product enhancer for object-based systems, and Sparks, a work-flow re-engineering tool with methodology. Coopers is currently offering day seminars for senior information technology executives entitled Executive decisions in object management. The brochure advertising the seminar has an aggressive message – to wit: there is an information crisis and the world has agreed that object-oriented technology is the way to solve it. According to Chris Stone, chairman of the Object Management Group, Coopers, Lybrand & Deloitte has been supportive of object-oriented technology for some time and despite the technology’s current low profile at user sites, management consultant for Coopers, Samit Khosla, is in no doubt that its time will come. He admits that at present he has only seen a trickle of blue-chip customers – six clients in the past year to be precise – but believes that this is largely because such clients are under duress from the recession and this is leading to short-term thinking. While Khosla knows that object technology is not a reality in terms of product yet, he does recommend that clients get started on a pilot project so that they can compare the time it takes to get the project up and running to the time it took their current system to become operable. Similarly, they should also make comparisons between the flexibility of an object-oriented approach and that of their current system. Khosla is a man who chooses his words carefully, and when he says object-oriented approach he is clear that being object-oriented is an approach and not simply a technology. It is not something that can be started by buying into a language and products, because that conception is part of the old model of computing whereby the data is separated from the process. And that is an artificial distinction introduced by computing that is no longer necessary with the object-oriented approach. Writing in C++ is not enough, it is not even object-oriented if the old data flow is kept intact. Conversely it is perfectly plausible in Khosla’s view to use an object-oriented approach encapsulating the data with the approach and code in a traditional language, such as Cobol. Encapsulation is the key to the new approach according to Khosla and it is being driven by pressure towards distributed computing where, by necessity, the distinction between data and process is being blurred. Khosla says that via encapsulation techniques, object technology can be used as a glue to bring together different technologies and that is why it is so important to systems integration. Khosla also has radical views concerning the way an object-oriented approach will change the way information technology is used within businesses. He believes that in the same way that data is no longer separate from process, so technology will no longer be a separate department, but managers will be in control of processes, each involving some technology.