FLOSS Weekly #27: Ward Cunningham

FLOSS Weekly 27: Ward Cunningham is, in my opinion, the best episode of FLOSS Weekly so far. Which I can say without guessing, because I listend to them all.

Part of my excitement is probably the fact that the things that Ward Cunningham has contributed are what I am pursuing to propagate at may dayjob right now as well as things I read a lot about for myself. I think what Randal Schwartz says in the end pinpoints it precisely what makes this episode that outstanding.

What an amazing conversation. It’s good when we get to talk to the legends, the people that set the platform that most of us work within. This has really been a privilege.

I can only recommend to listen to this episode as a whole, but here are some of the points that were most interesting to me.

• I wanted to have something where I could describe incomplete knowledge.

• Among regular Web pages, link destinations need to exist before they can be linked. This doesn’t fit incomplete knowledge.
• In creating a link, you point out to something that should exist. Not just for yourself, but for anyone using the wiki.
• Writing in a wiki is like programming. You start with the parts you know. When you know you want to call a routine, even one that may not exist yet, you create a link. Testing is seeing how it works. A document that’s just one quarter done can still be useful.
• Writing things down a little too simple than actually needed serves as an incentive to other people to contribute a sufficient solution.
• Some classic refactorings work better in static languages, but you need them much more often in these languages.
• Code that can be run early, even if partially incomplete, can be tested earlier.

• If you just do rails the way it’s intended, you’re doing agile without having a brain about it.

• On UML
  

  The notation doesn’t even allow the representation of the coolest objects I’ve seen in Smalltalk. If I can’t describe them, I am certainly not going to invent them.