Michael Watkins wrote:

>I will say that giving a new never-seen-the-framework before user a leg up in
>showing them how to structure an application beyond the demo-level of
>complexity would be a good thing, whether from a pre-built app layout or from
>tutorials or a script or whatever, to answer the odd "how do I structure an
>app" sort of question... questions that sometimes don't get asked.
>
>
That's a good idea. It would be nice if you could specify an alternate
skeleton when creating a new app; moderate to advanced users could have
custom layouts for specific jobs.

Dave Kuhlmann describes a Quixote skeleton at
http://www.rexx.com/~dkuhlman/quixote_skeleton.html. It may include
design choices that aren't suitable for a window-shopper (e.g.,
PostgreSQL is mentioned, though I'm not sure if it's used in the
skeleton; SQLite or Durus might make more sense) but you might find some
good ideas there.

I think that attracting new users is a different game, now that
TurboGears, Django, Subway et. al. are on the map. Quixote has always
been about good, solid plumbing, and that's extremely valuable; but the
average new user is probably looking for more help. I don't think it
would be a bad thing to have a TurboGears-like framework built on
Quixote -- not changing Quixote, just using it. But such a tool would
have to have some innovative component choices in order to stand out
from the pack.

Quixote doesn't need to be uber-popular in order to survive. Just as
many of us choose to use Medusa, in spite of Twisted being the current
poster child for Python networking, Quixote will maintain an audience
just by remaining simple and excellent. But it's realistic to expect
diminished growth in the Quixote community unless something sexy and
Quixote-powered draws attention back toward it.

Graham