On Fri, Aug 16, 2002 at 01:51:15PM +0200, Nicola Ken Barozzi wrote:
> Jeff Turner wrote:
> >I wrote the following blog entry about a day spent trying to use Forrest:
> >
> >http://www.webweavertech.com/jefft/weblog/archives/000027.html
> >
> >It was written just to let off steam at a wasted day, but I hope it will
> >be accepted here in the spirit of constructive criticism, 
> 
> Forrest is not yet ready to be used easily in other projects, because it 
> has not been really discussed and done yet.

Exactly. And why not? Because the focus has always been "let's improve
xml.apache.org", not "let's make a generally useful doc system".

I'm suggesting that the social engineering of Forrest is subtly but
significantly wrong. The fastest way to get a good-looking xml.apache.org
is *not* to simply advertise that fact and get cracking, as Forrest has
done. The goal is too abstract; not enough people care. Instead, one
should appeal to people's self-interest; broaden the focus, say "we're
here to make a brilliant doc system that anyone can use", and get people
involved. THEN, if the project has evolved correctly, the secondary goal
of a better xml.apache.org will be met.

Forrest micro-optimized on one tiny section of the problem-space.
Forrest hoped that a general solution would emerge from tackling a
specific problem, rather than tackling the general problem (with
consequently more resources) and letting a specific solution
(xml.apache.org) fall out.

As evidence for this view, I ask you again to compare Maven to Forrest.
Both have goals of unifying the websites of their respective
organizations (Jakarta and xml.apache.org). However with Maven, the
unification goal is subordinate; first, they want to produce something
cool and consequently widely used. In Forrest, the unification goal is
explicit, and as a result Forrest is only good at generating docs for
Forrest. Result? Despite Forrest's head-start, Maven has a much more
dynamic community, and Jakarta sites are a lot closer to being unified
than xml.apache.org sites.

I invite everyone to read 'The Rise of "Worse is Better"':

http://www.ai.mit.edu/docs/articles/good-news/subsection3.2.1.html


Practically, what I'm asking is that people take these thoughts to heart,
and if they agree, then start giving priority to problems limiting
adoption over problems limiting xml.apache.org coolness. The most
immediate problem is, I'm afraid to say, the forced Centipede
integration. By all means use Centipede to build Forrest, but don't force
it on users. To further crush Nicola's ego :) and to help spread adoption
of Forrest, I'll be spending tomorrow attempting a Forrest Maven plugin.


--Jeff

> 
> Eventually a new forrest Cent or an Ant task will get you going faster.
> 
> -- 
> Nicola Ken Barozzi                   nicolaken@apache.org
>             - verba volant, scripta manent -
>    (discussions get forgotten, just code remains)
> ---------------------------------------------------------------------
>