David Norfolk: Application Development Advisor
One of the constant characteristics of our "craft" is the worship of the new - we have the only serious engineering discipline where "if it ain't broke don't fix it" is accounted negative thinking. Luckily, DSDM has been around for about 10 years and, although it has been extended, I still recognise the basic principles from ten years ago - the process ain't broke and it isn't being unnecessarily fixed either.
However, new architectural initiatives for developers are always appearing - most recently, the Model Driven Architecture (MDA) and the Service Oriented Architecture (SOA) - and it is tempting to think that a new world is dawning. So, at Application Development Advisor we asked Tom Welsh (a Senior Consultant with Cutter Consortium's Enterprise Architecture Practice, Editor of Cutter Consortium's monthly Web Services Strategies, and an independent consultant and analyst specializing in middleware, object technology, and software engineering) to report on the real status of these initiatives. Tom has a wealth of experience going back to his days at Digital Equipment Corporation, where he was a hardware technician, software support specialist, corporate software developer, and senior technology consultant, before taking on the task of marketing Digital's object-oriented software products in the UK, and has seen development fashions come and go. I think it is fair to say that he is cautiously optimistic about these latest two but points out that they are very far from complete as yet.
In his article "MDA at the tipping point" (ADA Sept/Oct 2004, archived at http://www.appdevadvisor.co.uk/archive/index.php - free signup required), Tom concludes that "Today, the products that are supposed to implement MDA embody so many compromises and placeholders that it is doubtful whether they can be said to deliver the full MDA vision. Of course, we must walk before we can run. But that realisation only serves to underline the necessity of a reliable roadmap. And where is the MDA roadmap? It would be unreasonable to ask for one, because no single authority [not even the OMG] has control over MDA's future".
He quotes Dan Haywood's sardonic words, "Any technology that deals in meta-meta-models is going to be a hard sell, no matter how many analyst's reports have been written"; although he thinks that MDA does have a future because it is already being adopted in deadly earnest by organisations with a serious commitment to producing reliable, high-quality software quickly and, above all, making sure that it does what it is supposed to
In his follow-up article "Is SOA in hype(r)space?" (ADA Nov/Dec 2004, http://www.appdevadvisor.co.uk/features/index.html), he is even more concerned about SOA: "how has SOA, and the Web services model in particular, managed to attain such universal acceptance? There have been no big technical breakthroughs, apart from the decision to use XML as a standard message metaformat. The real advance has been a political one - IBM and Microsoft joined forces to promote Web services, and everyone else fell into line".
He notes that many vendor spokespeople have expressed similar sentiments about how SOA does not come in a box, is not a single product or technology, and so forth. A cynic might suggest, he says, that what SOA most closely resembles is a blank cheque. He notices inconsistencies in the SOA message: "We are told that SOA must be based on an enterprise-wide model and plan; but also that it should be undertaken a little a time, starting small. We are told that SOA will enable large-scale software reuse, as new projects can rely on existing services to supply part of their functionality; but there is no explanation of how responsibility and funding will be allocated, or how fixes and enhancements will be co-ordinated". He also points out that most of what the SOA advocates like to talk about has been possible for years or even decades - SOA is not yet a solution: it is, at present, a rather ill-defined set of requirements.
At ADA, we are generally neutral about the various industry initiatives that explode like supernovae in our in-tray - our readers can make their own minds up, if we give them the information in an entertaining manner. However, we do feel that it is important to present a reasoned alternative to the mindless optimism that is a feature of so many vendor presentations.
You can read Tom Welsh's articles on the web at www.appdevadvisor.co.uk or in ADA magazine (available free to developers in the UK, although we do ask you to sign up in order to read the whole archive).