Extreme Statements

He must remain nameless but he's neither one of the ThreeAmigos or one of TheThreeExtremos (as you can probably tell below). Nonetheless he's known worldwide for his books and other contributions to software development method since the early seventies. So last week I asked his opinion, in passing, about current trends in method, including UML and ExtremeProgramming:

UML - "oh I expect it will put the industry back thirty years"

Extreme Programming - "have you had experience of er ... Extreme Programming then? I see ... the initially sceptical programming team now say they wouldn't ever revert from PairProgramming and writing the UnitTests first. Well, I haven't read the book but I'm sure that programming in very small steps and the other ideas are great fun ... and work very well in the class of projects for which Extreme Programming is suitable, and not so well in the much larger class of projects where it obviously isn't."

For what it is worth, I am convinced that the first rather 'extreme' statement will now not be true .. precisely because ExtremeProgramming, with all its current limitations, has already broken the cosy but foolish consensus in software method described briefly in ModelingTrap. This in itself will ensure that UML will do far less damage than it would have otherwise.

How is XP achieving that? Through a mixture of ExtremeIconoclasm (the marketing) and the significant and substantial reduction in the DifferenceBetweenTheoryAndPractice that AlistairCockburn and other open minded methodologists are reporting (the grass roots confirmation).

But all this reminds me again that nobody has a monopoly in ExtremeStatements, whether true or false. Note that some very respected method people do make such comments about UML in personal conversation but have not yet made a single contribution to Wiki. Why is that, does anyone think?

I've already noted my concern that Wiki does tend to filter out strongly dissenting points of view, even from those with substantial experience, different perspectives and possibly great track records. I've tried to ask why in WikiSuccessCanInhibitNewWriters. -- RichardDrake

True. Anyone interested in Is UML Bad might read UmlAndBigDesign, UmlIsForPeople, or UmlTotalFailure, where the subject is discussed. The punchline seemed to be that it is "merely" a drawing convention, and subject to individuals' use of it. Actually, from my perspective, I'm happy to see few ExtremeStatements, because so few of them are defensible. On occasion I see someone "playing the part of" an extremist, to pull out the issues (and then usually confessing to their actual, mixed view of the matter). -- Alistair