A fictional substance that can be used to fix any problem.  Similar to a SilverBullet, except it has an escape clause:  if it fails, you obviously didn't use enough of it.

Referenced in an IBM television advertisement.  An engineer type is addressing executives in a boardroom, explaining that he's made a discovery that can instantly solve any server problems:  PixieDust.  Soy-based, biodegradable, easy to manufacture, and cheap.  The advertisement ends with a caption: "THERE IS NO PIXIE DUST".  There are, of course, IBM self-healing servers... sounds like pixie dust to me :)

Examples of PixieDust in software engineering:
	* OverTime
	* ExtremeProgrammingPractices
	* Outsourcing
	* ObjectOrientedDesign
	* HardWork
	* UpFrontDesign
	* Documentation
Remember - PixieDust '''doesn't work''', despite people thinking they should.
----
UpFrontDesign and documentation had been removed, but are now back on the list.  Many times, I've heard (of an ailing project), "Gee, it would have worked had you done more UpFrontDesign", or "If there'd been more documentation, you wouldn't be in the fix you're in."  The big difference between PixieDust and a SilverBullet is PixieDust's built-in escape clause:  if you used it, and it didn't work, then you obviously didn't use enough of it.

''As for XP practises and OOD, we'll just have to see...''

Anybody putting design and documentation on a list of "pixie dust" solutions isn't very professional. These things are ''requirements'' to making products in many, many fields of software engineering. In medical, aerospace, transportation, and most commercial automation you better have some up front design with documents to support it, else you simply ain't gots no product, period. FDA, FAA, DOT, UL, and any number of other alphabet soup regulatory agencies will assure this is so.
----
In anthropology this is characteristic of a ''ritual.'' In a ''ritual'' the procedures are so intricate, obscure and precise that you can't ever do it right. Because of the escape clause, the ''ritual'' is immune to empirical proof or refutation. Failure is redirected to the performer from the context, making ''rituals'' very attractive in management, methodology, or consulting. "You would have succeeded, had you simply correctly done what I told you to do ('directed' for managers, 'prescribed' for methodologies, 'conceived' for consultants, more or less.)" The trick is to recognize the dodge when on the receiving end of the ritual flavor of PixieDust.

The plan/execution tension is illustrated in the Bond book/movie ''From Russia With Love'' where Kronski's plan with Kleb's execution, fails. Pragmatically the purpose of a plan is that is be executable to produce the desired result. Thus, Blofeld makes the right call, eliminating Kronski vs. Kleb. At minimum, Kronski should have identified the risks, assumptions and preconditions required for the plan to succeed. Blofeld might have reasonably spared Kronski, had Kronski sought to learn from the failure or adjust the plan in progress. In stead, Kronski asserted that the plan was "perfect" declining to acknowledge his contribution to the results, and declining to engage in problem solving. The plan + the execution didn't work. That's everybody's problem. Blofeld's line: "We do not tolerate failure..." is also telling. The job of both Kronski and Kleb is to adjust plan and execution to get the job done. Kronski's deeper failure is declining to work the problem, instead defining his plan as a ''ritual'', placing all responsibility on Kleb. If Kleb is entirely responsible, Kronski's contribution is 0 and Blofeld is better off without the overhead.

Extrapolating this fable to the purveyors of software methodologies, selling PixieDust in the form of 'rituals'...
----
CategoryIdiom