One useful exercise when considering open source is to define your organizational personality with regard to risk. The more risk an IT department assumes, the greater the immaturity of the projects it can use. Supporting mission-critical applications with open source requires maintaining a larger pool of institutional skills.
The decision to use open source should be made carefully in the context of how critical a system is to business operations. The more important the system, the more skills are required to support an open source project.
The key question is how the demands for skill change as the risk goes up. Here are some categories of importance for systems:
If an experimental application breaks, it generally is not a big problem. Nobody expected it to work in the first place. When a system is installed for evaluation or on a trial basis to support one project, for example, a failure might not be a disaster.
A low-priority system can be down for a day without bothering anyone. This provides time to resolve problems, and life goes on without major interruptions in the case of an outage. A Wiki used for documentation or project management, or a weblog used for internal use, can fall into this category.
Operational systems must run well and not be down for more than an hour or two. This is where the required understanding of an open source program and the skill level start to creep out of the intermediate level and toward the ...