Is an Agile methodology going to stand up in court?
Bill Anderson reports that April’s meeting of New York City SPIN (that’s the Software Process Improvement Network) had a talk by Ed Yourdon about software litigation. Software litigation, Yourdon reported, is increasing, and as someone who’s often involved in these cases, he’s increasingly running into lawyers with backgrounds in engineering and computer science.
In lieu of a Powerpoint, there’s some sort of mind-map dump available which gives the gist of Yourdon’s talk. One telling note is under “Processes”: Beware XP, RAD, “agile”. This is a bummer, but not really a surprise. Large corporate software processes often produce massive documents, and one explanation for this is that people involved are trying to cushion the fall when they fail. When your project fails, you’re less likely to be fired at some companies if you can point to a massive spec doc as proof of your efforts to control the project.
The same risk aversion would easily be built into any legal proceedings. While it’s easy to imagine using massive documents to explain away failure to an inexperienced judge or jury, Agile methodologies would probably be a much tougher sell. Such methodologies, with their emphasis on human relationships and close communications as opposed to writing everything down to cover your ass, seem sort of tailor-made to fail in court. (What that says in general about the effect on society of explicit laws, as opposed to implicit human relationships, is left as an exercise for the budding anarchist reader.)
If such litigation continues to grow, we have to be concerned about law setting industry best practices for us—in an industry which is far too young to have any best practices at all. Computer programming isn’t like medicine or civil engineering; it’s only a couple of decades old, and we don’t really know what we’re doing, and the only way to improve is to keep trying new things. But if civil court cases ending up implicitly telling software vendors “you will be sued for these practices, but not these”, I can’t imagine that as a good thing for the field.