Propagate Xp

See also PropagatingXp

Based on discussions in ExtremeProgrammingInEnemyTerritory it seems difficult to get the chance to do a large scale ExtremeProgramming project. With that in mind, what is it going to take to PropagateXp?

It'd help if Beck, Auer, Cunningham, Fowler, and Jeffries would finish "Playing to Win".

That's a very small part of it, but thanks :-)

There is plenty of information out there for folks to "behave extreme". It will take a project that simply can't be done any other way. It will take communicating with the project sponsors on the customer side what XP can do. It will take communicating to the programmers how to get started doing it.

It's not that small a part of it, in my view. You have a small but dedicated cult here, plenty of people (including myself), who pitch XP at every appropriate opportunity, and the occasional inappropriate one. Line programmers will buy in to an experiment because anything is better than what they have. Team leads will buy in because they remember the line. Project managers can usually afford 3 to 6 weeks if they think their lead wants to try. No-one in middle-management will buy in because they have neither understanding nor a cover-your-ass position. They don't even have the three books I bought them because there aren't any damned books I can buy. I cannot close the deal for an XP experiment on the basis of 'some guys I know on the net did a project called C3'. I cannot close an XP experiment deal on the basis of some web citations. So I work perpetually underground, without the knowledge or support of a project manager. PairProgramming, allegedly fundamental to XP, is virtually impossible to achieve in this environment.

Guys, you will not capture the market without publications, and plenty of them. Please don't hesitate to contact me if there is ever anything I can do to help on this score, from copy-editing to typing to ghostwriting to article-hawking to midnight counseling to a quiet place in the blue ridge mountains for thinking to more beer to... you get the idea. -- MichaelHill

Here are some other thoughts.

Illustrations in C or C++, the standard industry environments. Smalltalk is harder to sell than XP, and illustrations in it are difficult for weak programmers to read. Java is only a little better.
One successful desktop-based shrinkwrap packs more punch than a dozen large commercial in-house products. It easily leads to How-We-Did-It articles all over the industry rags.
An iteration-by-iteration case study of a single six-to-twelve month project that started with well-targeted requirements would do wonders.
Sharpen the distinction between XP and CowboyCoding. Otherwise you cannot....
Criticize the behomoth that the three banditos have created. Without that criticism armchair programmers, which is most middle management, will not be able to take the 'risk'.
ExtremeProgramming makes it sound like we're surfers, which hardly helps to convince people we're not cowboys. Lose the name. (reNameXp).

Soon it will be time to look backwards and see who starts to trail. How did C++ beat Smalltalk? It gave people 80% of what they thought that they needed before they knew the difference. The same thing could happen to XP. Another process could come along which is not quite as extreme, promise a small improvement over traditional techniques without turning the world upside down, and deliver. It is hard to stage a revolution when evolution works.

What would an XpLite look like? Some BigDesignUpFront + UnitTests + ReFactoring and tighter iterations. That is a stable configuration and it allows people to back out of corners. It is an easier pill to swallow and it is something that XP will have to beat or embrace if it ascends.

To me, the best thing that the XP authors can do in their publications is draw a map that shows how to get incremental improvement from wherever you are and move towards complete XP. Ward's skiing analogy for XP points out that you get most of the bang for the last N percent, but people have to see how to get there and see how they are getting better. They need to know what the next step is along the path. This means that XP would have to lay claim to the path that is made by incremental adoption of ExtremeProgrammingPractices and ExtremeValues. If it is seen as a pot of gold at the end of the rainbow, people who do not delight in the journey may never attempt it. IMHO, XP should not be just the ideal, but also the path. It has to stake a bigger claim. -- MichaelFeathers (but that's just my opinion, I could be wrong).

See AdoptingXpPatternLanguage, CrossingTheXpChasm

CategoryAdoptingXp