I gave a presentation at the IBM Mashup Summit a couple of weeks ago, and received in return a number of positive reviews. I thought I would turn it into a series of posts that go a little deeper into the topic, which for the vast majority of the market is still pretty new stuff.
This first post will take on the word "mashup" which even today causes snickers in enterprise circles. The fact of the matter is that it’s just a word, whether you call them composite applications, web services orchestration, SOA, next generation EAI, orÂ mashups, it just shouldn’t matter.
I can tell you from personal experience that if you are in front of a venture investor you won’t get but a few minutes past uttering the word "composite" before watching eyes glaze over. Likewise, if you are in front of enterprise IT people and mention "composite" you will quickly turn friends into foes as they silently recall all the previous attempts to sell them composite software that only resulted in a lot of money leaving their budget for a lot of technology upgrades they are still struggling to get implemented.
IT people know what composite apps are and generally think they are a good idea, but they also put their defenses up because everyone that has stepped into their office and tried to sell them a composite app framework was trying to sell them a future.
So in the end it’s not that "mashups" is a great term, it’s that the other terms are so negative. In an industry that regularly takes really cool product code names and turns them into awful product names that then get distilled down into acronyms, well I really don’t mind the term mashup.
Mashups have a magical quality to them in the eyes of users who see them as emotional more than technical. While we may look at this as reusable software componentry loosely coupled to achieve lightweight universal EAI, consumers (remember, business users don’t stop being consumers when they go to work) look at them and say "great, this works the way I do!". It’s really that simple, an emotional response keyed off giving users more control in an increasingly complex world.
I don’t do business process, I do tasks and when a group of my coworkers gets together to organize tasks into an activity, well it’s a business practice. The idea that I can take services I am already using and mash them up into an application that can make my life a little simpler is really appealing to me.
In the posts ahead I’ll look at user interface issues, security, enterprise service buses, the API business model, and performance. The bottom line from my experiences over the last year is that this mashup thing is real, people want them and enterprise IT pros know that this is one way that they can unlock a lot of value in the big applications they are already paying for. It’s a complex topic and once you get below the surface you find a lot of big issues that have to be overcome and a potentially meaningful realignment of the software industry as a consequence.