The promise of SOA - what has it brought so far?

SOA has been around for a few years now, time to see what it has brought us. Especially since SOA 2.0 seems to be an excuse to us all to forget the imperfections of SOA and hurry over to the next hype, leaving SOA behind in its own rubble

I think SOA is a good concept. IT is bad enough for the business as it is, with all its inherent diversity: one business, a few processes, and dozens or hundreds of applications supporting it. And, since CRM and ERP, we all know very well that the one-size-fits-all promise is yet another Utopia, slowing down your business rather than doing the opposite

Doing SOA the right way

Here are my 4 points:

1. Understand the diversity, and the evolutionary law: everything will always change. You might think that you achieved a lot by homogenising today's mess, but just wait 5 years and then have another look. Thou must always translate - and leaveth that to professionals. There is no one-size-fits-all

2. Diversity as in applications, but also as in languages spoken. In Europe, 2,000 years ago Greek and Latin were cool, then vulgar Latin took over, French was okay since 1600, English rules since 1900.

Architecture as a pressure cooker

Right about now I'm reading about a company that wants an enterprise-wide Service Bus "to glue it all together" (my own words) and create an agile, flexible service oriented architecture (their words)

Now, how well have they thought about the different departments, each dealing with different business pieces, consumer sub-markets, various government regulations, different timings (8/5, 12/6, 24/7) etc?

Not at all. It's all tech-talk. Here's the worst part: The ESB must support SOAP over JMS outside, and SOAP over HTTP inside. Described in WSDL. Now I know exactly what that implies, but the authors probably don't have a clue themselves.

The sudden death of EAI

After decades of more or less succesfully integrating databases and applications, there was a proven model: hub-and-spoke architecture with a canonical model

It was pretty perfect: the evolutionary as-is and to-be diversity of any IT-landscape was absorbed by an almighty interpreter-translator in the middle, who would take all the different languages and dialects and translate them to an intermediate language that would be that enterprise's business language
Just as in the European Parliament, where they call this a 'relay' language. Now, if anyone knows anything about levelling (language) boundaries, it is right there. They handle 23 different official languages

In 2010, Twitter will be the pulse of the planet

It's the end of year, a time of looking back, and ahead. A fun time to make predictions, and look back at predictions made earlier - although that usually is much less fun

I predict that everyone will have a Twitter account in 2010. Every company will also have one, and use it too. There will be Twitter boards in public places, public ones as well as private ones, some of which will be censored to a degree, much like the delay already present on US radio- and TV shows.
The private Tweet boards will also monitor Foursquare and BrightKite in order to see what the world is thinking about that particular place

On The Acquisition Of Knowledge 2/2

In my last blog post I talked about what to do with the information you get. This will be about how to get information

There's data, and there's information
Search engines like Google search and aggregate data thereby enabling it to become information, which is invaluable, but not too easy. Googling is an art, and requires skill
Human "engines" such as Wikipedia explicitly present information in the best possible and moderated way. Wiki is an absolute fantastic source of information for everyone in the entire world (I just can't stress that enough) where even the hottest debates are held