There is a LOT of hoopla lately over an article proclaiming that SOA is Dead. Of course, the title is attention-grabbing and not an accurate representation of the article’s content (the author even suggests that “SOA” is dead, but not “service-oriented architecture”). None-the-less, you can’t have it both ways; you can’t label something as an “obituary” and then go on to write how the right arm will still work if you prop it up at 45 degrees and twist it three times. So, I will simply state that the author is wrong and suggest a more accurate picture of the state of current day SOA.
SOA projects are expensive and unpredictable. That hasn’t changed and that is precisely the “problem” with SOA. And one can argue that the expensive SOA project has, indeed, died. However, SOA is not the only traditionally expensive approach left in the market today. I find that many of our open-source projects are completed at costs much lower than an enterprise solution would have required. However, I also find that we complete projects in general at a much lower cost than most other organizations.
So, the point here is that organizations will begin to get much smarter about how they spend their money. I don’t think anyone can argue with that. Has SOA matured to the point where there are enough experts and experience in the market that SOA can be easily done in a cheap and reproducible manner? No. When Java first came out, did enterprises take quite a lot of time to implement solutions/architectures cheaply using it? Yes, and I would argue that most organizations still spend too much on Java-based implementations.
The fact is, given smart-enough people at low-enough costs, SOA can be implemented in a cheap-enough manner even for today’s economic condition, while still delivering the benefits it promises. There are wrong, expensive ways of leveraging any technology, and there always will be. We can all welcome with cheers that successful companies have no choice but to limit doing things the wrong, expensive way as they get smarter and more efficient.
So, if I were writing a similar story of the economic impact on today’s technology, my article would have been “Unpredictable, Expensive Solutions are Dead” (where SOA happens to get thrown into that bucket by many companies who have not been able to implement successfully-enough).