Velvet Underground

Chapter 19 of KentBeck's book "eXtremeProgramming eXplained: EMBRACE CHANGE" describes the process for adopting ExtremeProgramming (XP) as follows:

 1. Pick your worst problem.
 2. Solve it (the XP way).
 3. When it's no longer your worst problem, repeat.

Like most of XP, the approach is nice and simple - almost common sense. Indeed a lot of people instinctively have this sort of approach. I would share with you the true story of the Velvet Underground. A story in which the approach of explicitly addressing the worst problem made the miserable lives of a handful of developers tolerable and eventually rewarding.

The story starts in the late 80's when I changed jobs. The company I joined were (and probably still are) a small, successful, software development company. They were successful because they had spotted and entered a niche market. They were successful because the company's founders were real experts in that market. The success was very hard to earn because they didn't have a clue how to develop software.

When I joined the company it soon became apparent to me that it was a miracle that they shipped any working software at all In fact most of the software they shipped required several modifications before it would work. At that time I thought I new how to develop software and set about trying to change things (I still think I know how to develop software but since my opinions then and now are so different they are probably both wrong). After 3 months I had made no progress. Most of my time was spent fire fighting. Any suggestions made for improvement where rebuked or ignored. I was told that such things weren't my concern, I was told to stop causing trouble, I was told to just do my job.

Then one evening, after a particularly awful day, I went to a local pub with two of my colleagues. We talked about work. We complained about work. We talked about how easy it would be to improve things. We complained about how we were not allowed to improve things.

On the way home it occurred to me that I now had two allies in my quest for improvement. The following night I invited my colleagues to the pub again - I had a plan.

During that second long evening in the pub we made a long list (and I mean a long list) of things wrong with the company. We then placed the problems in order with the worst problems at the top of the list. We resolved to tackle these problems one at a time. Some of the problems were just to big to address initially but these were not removed from the list. We just started with the biggest tangible problem.

We agreed to meet each week to discuss progress, the next problems, and how to address them. We were worried that if certain members of the company found out about what we were doing we might not be popular and there might be attempts to stifle what we were doing. Rather than take the risk we decided to keep our extra curricular activities a secret. For a reason I can't remember we decided to call ourselves the Velvet Underground.

The first problem we addressed was to get the development PC's networked. Things were THAT bad! The next biggest improvement came from automating a good deal of the testing. Another big problem was preventing the Managing Director having unlimited access to source code - I have to confess, this took some blue language but earned me a lot of respect.

After 3 months things had improved to the extent that we no longer felt we had to meet in secret. Indeed discussions about improvements were a regular and open. Although I cannot remember ever admitting to the secret meetings of the 'Velvet Underground'.

---PhranRyder

Also, a very influential band...