Scott Ambler has started another Agile Adoption survey with Dr. Dobb's, the 2008 version this time. Though I applaud the fact that experts are trying to find out if and how agile methods work, I have a few problems with the survey questions:
I am asked which of my projects has been agile. But what is an agile project? Is it agile when we use iterations? When we use stand-up meetings? When we apply TDD? When we apply timeboxes? Is it agile when we throw half of the workstations out of the window so that developers need to program in pairs? Any of these? All of these? I don't know. Can someone give me a binary definition of what is agile, and (even more relevant) what isn't?
I am also asked how big my projects are. But what is a project? Is it all the effort put into version 1.0 of a new system? Does it include version 1.1 shortly released after that? Or was that a new project? Does the project include all iterations and releases over a timespan of several years, until work on the product died together with the mental collapse of our customer? Can someone give me a binary definition of what is a project, and what isn't?
To be honest, I am not able to answer those two questions. And I don't believe other people can. And if they claim they can, they will definately give me very different answers, so I will not be able to compare their survey data. Therefore, even though I will be sure to check out Scott's results and evaluations, I won't be able to tell if they actually mean anything.