Why Standards and Open Source Will Save You Money
You’ve probably seen a commercial at some point with some guy throwing money in the air, declaring his intent to “Save you money!” Of course, the deal invariably required that you pay money to “save” money, which makes marginal sense at best. Still, this perennial exuberant exclamation has served as a regular reminder of the natural human desire to pay less than retail price (whatever that means).
Saving money these days carries vastly more weight than it has in previous years – and that trend will surely continue throughout 2010, one way or another. The reason is simply because lots of money really did just disappear, and new money will take time to infiltrate the system (read: your pocket). So, what's an information manager to do in the meantime?
For all kinds of reasons, go open source. In the first place, you have that wonderful transparency of code. As just about every open source pro will tell you, the really magical thing about the stuff is that “bad code goes away.” That's quite good, actually, because bad code is like a bad gene: It can rear its ugly head at any time, in any way, for any reason; which makes troubleshooting downright painful.
Furthermore, open source will save you money in at least two ways. One strategy is to dabble with it, then tell your enterprise software rep that you're having some success with the stuff. “At first, the developers weren't sure about the environment, but then they got the hang of it, and things are kinda workin' out now...” That kind of language might just net you a lower maintenance fee, or some wine.
You can also save money by scaling back on seats, or nixing whole packages, for various projects. On a recent episode of DM Radio, Forrester Analyst Jeffrey Hammond noted that open source software has definitely arrived in the enterprise. In fact, his testimony rings louder with every listen. During the radio show, he discussed some surveys Forrester has done recently on open source adoption throughout the IT stack.
After listing a litany of robust percentages for developers using open source technologies, he stated: “Overall – and I think this is the key point – only one in five developers have said that they have not used open source as part of their daily development activities, or things that they've used to deploy an application. So, at 80 percent, you could argue that open source is a de facto standard for IT.”
Four out of five? Only one developer in five has not used open source?
Your humble DM Radio host uses open source technology every week. The application that encodes our audio stream, SAM Broadcaster, leverages MySQL as its database. (Yours truly thus knows a little bit about configuring a database instance ... which can go wrong for any number of reasons that may or may not be associated with whatever might be happening somewhere on your computer at that time.)
Where were we?
Oh, yes, saving money! As Yves de Montcheuil of Talend told us during the show “Nothing is free. Even if you can get a license of the software at no cost ... there will always be a cost of implementation, a cost of development, a cost of deployment, buying the servers and so forth.” There will always be some costs associated with setting the stuff up, managing its use over time and adding functionality.
And then there's the data ... yes, the actual data, the stuff that business intelligence dreams are made of. What about that data? How can standards be used to save money?
As Loretta Smith of DAMA International told us, the data really is the key, and it should be managed very carefully over time. “The interesting thing about data is its lifecycle is much, much longer than application code's lifecycle. So when you develop data structures, they typically outlive the original application that they were built for by at least five to 10 years.”
Ten years? At least! Some legacy systems contain data structures that were designed 40 or 50 years ago, often still running on old, if venerable mainframes. (Some will tell you that the old iron is the best iron in many ways — if not for the energy efficiency issue).
Smith continued: “Frequently, the largest portion of a project's budget has to do with moving legacy data into a new system. And so, the more you apply your standards to the development of the data structures that are going to support your applications, the more you can restrict the friction for the flow of data, either within the systems in your organization, or even across international boundaries.”
This is where standards really shine, and what master data management is all about. By employing standards at the semantic level, with a corporate taxonomy, for example, a company can manage its metadata in a most meaningful way, and that will make for much happier – and less expensive – times down the road when new systems come online.
“The reason that many people in technology have not paid attention to it is: we tend to focus on the development process, not the ongoing maintenance and production use of things,” said Smith. “So I think the business is suffering from the short-term perspective that many application projects have had towards the development of data.”
For the rest of this story, please go to http://bit.ly/OpenSourceDMRadio.