Perspective: Sticky tape holding back the revolution

So component-based development is about to take over the software industry, according to researcher Ovum. By the year 2002, Ovum?s latest report says, reusable components will form the basis of 60% of enterprise software development.

This is bullish indeed, as the software world has been waiting for decades ? 30 years, according to Ovum ? for component technology to become a dominant force.

The potential benefits from some kind of component approach are enormous. If re-usable modules or components are used, application development will be much quicker and the applications themselves more reliable.

Above all, software bec-omes easier to maintain. If a component needs to be upgraded, it can be swapped out and replaced with another. No more sifting through millions of lines of tangled code trying to find out the implications of a change in some core routine.

Components promise a controlled world of ordered software implementation. A rules-based world that relies more on discipline and logic in design than on the isolated ingenuity of a few star programmers.

Some cynics argue this is cloud cuckoo land. The IT world has known for years that there are much more efficient ways of producing software applications than the hotchpotch of different techniques, methodologies and languages that are thrown in at the development process.

But object technology hasn?t really taken off yet, so why should it happen now?

Ovum argues that the success of Microsoft?s Common Object Model (Com) and Javasoft?s JavaBeans technology has paved the way for components. As organisations struggle to exploit Internet technology, they need to develop applications that communicate information from legacy systems to Web servers.

In the longer term, Ovum sees the Object Management Group?s Common Object Request Broker (Corba) standard prevailing as the architectural tool of choice, and Com becoming the standard for implementing components on the desktop.

If the world could be this simple, then Ovum might be right. But it is not.

The IT industry is up to its neck dealing with year 2000 and the euro.

One section of the IT industry argues first that we would not be in this mess if applications had been written using components, and second that as applications are being investigated, fixed and updated as never before, now is the time to think about a new, component architecture.

While some organisations are working hard to implement next-generation architectures, most have adopted a ?sellotape and paste? approach to both year 2000 and the euro.

I?ve been an object evangelist for some years now, but I see a lot of fixing and patching, making do and struggling through over the next five years.

Sadly, the component revolution will have to wait.