Magic Slurpies

2004-09-01

In the continuing conversation about what's wrong in the Jabber community, Peter Millard observes:

There's much truth here, I think. In a follow-up IM session, Peter further observed that in order to enable people to "scratch their itches", it's necessary for those working on existing projects to:

  1. provide code that is hackable
  2. provide enough info about the code (where to find it, how to dive in, etc.)
  3. provide code that is modular enough so that people CAN jump in and contribute

In other words, we don't have more contributors because the barriers to entry are too high and most coders don't have the time or energy to grok an entire, undocumented codebase before contributing. A further reason for the lack of contributors is that most Jabber projects (and probably most open-source projects) don't have well-defined roadmaps, so Joe Random Hacker doesn't have a good sense for what needs to be done and where he can most productively contribute.

I once worked for a guy who did some projects for the 7-11 chain of convenience stores, and he used to say "There are no magic slurpies." Translation: there is no big drink that you can gulp down and everything will be fine. Solving these problems requires hard work, one project at a time. Projects that make it easy for other coders to contribute will thrive -- those that don't, won't. It's as simple as that. But well-designed, hackable, modular, fully-documented code doesn't grow on trees. It's hard work, and we'd be well served to dig in rather than just talking about it.

Make that mantra #3: less yammering, more code!


Peter Saint-Andre > Journal