With respect to the "style" of FAQ lists, and also in light of an
earlier discussion we were having about the distinction between
"volunteer" and "commercial" FAQ lists, I'd like to bring up a
completely different distinction: what we might call conventional
vs. contrived FAQ lists. Here are a few paragraphs I wrote on
the subject last fall, in the introduction to my review of a
manuscript (which happened to be of an FAQ book, of more the
latter kind).
[snip]
The "traditional" approach, if it can be successfully
accomplished in a given instance, has a distinct
advantage: the questions it covers have some real chance
of matching the questions which a reader is trying to
answer, even if the reader has tried and failed to find
an answer using more traditional references. In fact,
the process of identifying questions to be answered in an
FAQ list by noticing which questions are asked in an open
forum automatically discovers those questions which the
traditional references, for whatever reason, do *not*
answer.
[Al, here]
...or don't make sufficiently prominent. Lots of our FAQs in
Lynx land are answered in the documents, if you can ever figure
out how to find it... There is a generic problem in software
that the people who know the answers don't understand the
coordinates of question-space.
Well put, Steve. The traditional FAQ is based on listening to
the customer. That is one of the foundations that made
news.anwers and RTFM such a great invention. What it gathers is
not just any answers. It collects the answers to the questions
people need answered.
-- Al Gilman