For small values of B

Hey! My post on the Aardvark spec was quoted in The Farm - way to make me feel important...

I should clarify my position by pointing out that I don't see any value at all in trying to design every little facet of your software before you start. That's classic, old school, kidding yourself project management, and you're bound to fail, because reality will kick you in the ass everytime.

What you should do, is design a detailed central specification to describe your common architecture and your overall feature set - one that lets every member of the team develop a shared vision of what the product will be, before it exists. One that lends itself to a believable schedule. Armed with a clear idea of what it is you're doing, and a timeframe people don't think came from a magic eight ball, and you're ready to get started.

Seems to me that was what Joel's spec was all about.


(Whether that's agile, traditional or regurgitated-weasel is kind of unimportant.)