Mashup Recycling: Now this is Green IT!

Sometimes ideas come from the most unlikely of places. Yesterday I got off an elevator and wondered if it was one of the energy-saving models that regenerates power. They ‘give back’, in a sense, making the next round trip less costly than the previous one. An instant later I found myself noodling this very same principle in the world of mashups.

Mashups can be used to build new mashups. They make the next mashup easier to create. It this is the very quality that gives mashups the majority of their ROI and separates 'once-and-done' custom mashups from long-term systematic mashup solutions that achieve a kind of fission via a community of mashers. And in my considered opinion, mashup reuse is one of the most important yet under-utilized qualities of mashups. I think mashup reuse needs it’s day in the sun.

Enterprise mashups can be built on raw data sources (SQL databases, for example) and more formal services (like REST and WSDL). But mashups can also be built upon other mashups, assuming your mashup tool lets you publish the mashup with some sort of callable service interface. So more mashups mean more raw materials to go into the next mashup, creating a kind of network effort. You’d think this might be a no-brainer but I frequently see mashup implementations that that have no reuse at all.

Most mashup vendors show a mashup lifecyle as a circle; the last arrow that closes the circle is the 'reuse'. But this kind of network effect does not happen automatically. Much like the best SOAs, mashup reuse requires some kind of infrastructure to encourage reuse, such as a mashup hub/repository/registry.

Much like the worst SOAs, it is easy to forget about reuse and end up with a bunch of point solutions that stand alone (and miss a lot of the real ROI value of the technology). I see lots of cool mashups that are custom-coded in Java or Javascript. Most of these essentially become a 'mashup silo', with no option for reuse of that mashup's service in the next mashup, particularly for mashers makers other than the original creator.

But technology alone doesn't equal mashup reuse. A second common stumbling block is a social one. You need to change mashup making behavior. Sure, most enterprises have a common organizational mission but that doesn't imply that these same organizations motivate their employees to explicitly think in 'reuse' terms. There’s no magic bullet in any software package that will do that. Mashup reuse takes knowledgeable, committed leadership that can take independent-minded people and make them into a community of mashup makers that look for reusable mashup inputs and strive to create reusable mashup outputs.

Much like real recycling efforts, you need a pro-active, systematic, persistent ecosystem-wide commitment for real mashup reuse. Keep it green, baby.