Complexity or completeness?

I was asked again this morning about complexity, in relation to my view on both hardware and software. It would all be so simple if we were a start-up provided we gave you the “Power to leave” you could have it our way, or no way.

When I got back to my desk I went looking for a blog comment I wrote on complexity. For completeness and because it came up this morning, here is my response.

“The real challenge though that IBM faces, is not the complexity of our products, but the complexity of our customers.

If we were &Ampersand. small software company period, or an organisation we could do just a single product and say “there, thats SOA/ESB” its great like it or lump it.

However, that wouldn’t be much use for the millions of customers over dozens of OS’s, and four hardware platforms, built up over 30-years, who want to embrace SOA. Sure, many of them can and will do it without our help. Heck, some of them even do it without our products ;-( but generally while we have often intimate knowledge and understanding of their systems, they still want a shopping list of options rather than just do what we say.

So, that leads not to complexity, but rather to completeness. Many products with interfaces to, and programability for services based applications and infrastructure.

As always, people would like a single message, a single voice, but mostly customers don’t want a single product unless it’s the one they currently heavily exploiting. Even then they want something else to integrate to it, with it, or from it.

This is why open is key. Embracing web services, getting involved and implementing WSRF, WSDM et al. will pay off in the mid-term for both the customers and for IBM. The ability to implement applications around a services base, with a strong mediation engine, that participates in and can support a robust set of open industry standards is key.”