Digital Mileage

Academia calls it “publish or perish,” meaning one’s worth is proven by sharing knowledge with the world. Saying you know how to solve a problem is easy enough, but can you prove it? Can you show it?

In publishing their first white paper, Cobalt has done exactly that. A white paper is a document that addresses a specific issue and discusses how it can be solved. Jason Taylor, advisory engineer for product development at Cobalt, summarizes why he wrote, “Solving the White Label Problem:”

The White Label Problem is the ad hoc customization syndrome that software service operators suffer. It allows other vendors to resell the “white label” services as their own branded products. This problem usually happens when you’ve been successful enough at your offering to attract some large, important customers. These customers often have high expectations of quality and delivery efficiency, a pressure that is, in turn, applied to the White Label operator’s system; and it can have ruinous effects, if done without some kind of unifying design or pattern.

When I sat down to write, “Solving the White Label Problem,” I really thought it would be pretty simple and short. I still think the concept is simple: complex customization requests from large, important customers who resell a system’s services apply pressure to the system. This pressure is an under-appreciated, perhaps unidentified, aspect of software engineering.

What I found in writing the white paper is that even simple ideas are hard to fully describe and build to the level of a thesis. I developed a running system, which generates the visuals used in the paper. I had to tune the code and be very careful about unnecessary logic. Describing the concept of multi-dimensional code paths was difficult, even though it seems like a simple idea. In short, I didn’t really know what I was getting into.

Systemic, fully isolated customization of behavior will become a “hot” area of software engineering research; because I’ve seen how important it is at Cobalt and in other systems. The solution we have at Cobalt has sustained us during a period of explosive growth in the frequency and complexity of customizations to our system.

We and other firms would benefit if framework creators considered this aspect of software engineering when designing their systems. Another benefit would be if creators answered the question, “how will this framework generally support radical, dynamically-applied variations of behavior?” I’m hoping that some of the people who read this white paper will acquire the vocabulary and tools necessary for answering that question.

Cobalt’s needs, along with several levels and types of programmatic control, may be somewhat extreme; but there will be more and more software service vendors as industry trends continue toward infrastructure commoditization and software consumer usage sophistication.

Cobalt spends time and effort finding creative solutions to the problems that both the automotive industry and technology professionals face today. Cobalt’s unique place in the industry encourages innovation. Cobalt’s first white paper is proof of that innovative spirit.

“Solving the White Label Problem” will be presented this week at TheServerSide Java Symposium in Las Vegas and can be download at http://code.google.com/p/thewhitelabelproblem/.

The Cobalt Group

www.cobalt.com
www.twitter.com/CobaltMarketing
www.youtube.com/cobaltlive