Have you ever noticed that the more likely a given Open Source project's
community is to tell you to RTFM, the less the chance is that they have
a FM that's worth R'ing? I don't think that this is a coincidence.

This is a book about best practices in Open Source documentation gleaned
from the last 15 years of experience.