Saturday, May 20, 2006

Objects in the mirror may appear more Open Sourced than they actually are

Over the last five years, I've done a few projects on Open Source, first at a strategy consultant firm and then for an enterprise software company. I've looked at Open Source from three primary frameworks: trying to describe it as an ecosystem, dissecting it as a business model and detailing it from a customer perspective. From these projects, its apparent that open source is not a monolith, it's different things to different people. This comment raises the hackles of some who object to the idea that Open Source is a "flavor" that can be injected into most anything a company does, not necessarily requiring an open source license.

Yes, I agree from a pure taxonomy standpoint, many of these things might not be "purely" Open Source. However, the Open Source label really doesn't mean much for a business or customer. Business or customers are after the results that open source can deliver to them, not open source itself.

Why this set-up? Because things aren't always what they seem. There are a lot of new companies out there branding themselves open source, but when you dissect their business model, they are really quite traditional. On the other hand, there are businesses out their that don't appear to embrace open source, but when you dissect their business models, they have surprising open source "flavor". If you heard there was a development community that had 500,000 members that contributed 80% of the content, you might think it was developed around an open source company? This actually referring to SAP's own development forum (SDN).

There is still a place to talk about open source as a classification. However, as a business or customer, the focus should be on the resulting benefits of open source. Business and customers are focused on development cost, quality, product cost, distribution, customization services, innovation and community support. In some cases, open source might be the best option, while in other cases traditional models are the right choice. The bottom line, though, is look beyond the models to what is actually delivered.

No comments: