Over at Code Craft Kevin is arguing that agile methodologies make for more code, rather than less.
It’s an interesting idea — that the waterfall model introduces some sort of Darwin-esque survival of the fittest trials that only worthy features/products can make it through.
I’d argue the opposite — that, if done right, agile methodogolies (esp. when combined with dynamic languages) bring about less code. And that the agile process, especially if it’s augmented by user monitoring/metrics, creates an actual Darwin-esque feature ecosystem.
If you can write a feature in less code (with Python, Ruby, PHP, your dynamic language here), in less time and deploy to users quickly — then your code and business processes are much closer to the “pulse” of your userbase.
On the other hand, a long waterfall process probably errs on the side of the personal tastes and preferences of the managers and stakeholders in the review process. If that’s mated with a more-code oriented shop then you’re dangerously removed from that “pulse.”
Meanwhile… back at Creating Passionate Users they’re asking the question what
comes after agile in the evolution of development processes.
Right now agile/xp seems like the be-all, end-all, as fast and as good a process as you could ask for. But then I also thought doing pointer arithmetic and remembering the semi-colon was fun.
Change is constant, there’ll be something better… let the battles begin.
Will my Mom ever grok RSS?
I’m not strongly inclined to say it’ll never go mainstream, but I am surprised that it hasn’t gone more mainstream already.
Will RSS only ever be the de-facto simple REST API for content providers?
I think this one’s still in play, for now, it’s an interesting idea that RSS may be a primarily power user- and developer-focused technology…