idea webs

[Jim Little]

From: "Tril" <tril@tunes.org>
> In general what features do you look for in a free-form associative web ?

Wiki remains the best I've seen.  There's a Java-based version that uses
WebMacro (http://www.webmacro.org) -- it seems fairly complete, and it
supports versioning, which seems like it would resolve the vandalism
problem.

Also, a note that will undoubtably be unpopular.  Tunes doesn't need more
documentation: it needs prototypes.  Water deserves kudos for being the only
one still involved with this project to produce working code.  Why not focus
efforts on that, rather than yet another pass at the website?  Or, if you
don't understand or agree with Arrow, produce something else.

When I was working on Prism, I found myself updating documents over and over
as a way to procrastinate.  I had set myself some hard tasks, and couldn't
figure out how to solve them.  So instead, I wrote document after document,
and came up with a lot of "5 year plans" that allowed me to dither around at
the edges rather than tackling the heart of the problem.  Eventually, the
project died because I realized I had spent years accomplishing nothing of
value.

In commercial software development, you want to deliver working software as
soon as possible so that the customer continues to fund development.  The
most effective way to do so is to create a complete working system that only
implements a few features, then fill it out and refactor over time.  Rather
than deliver a perfect piece of software in twelve months, you deliver
imperfection in one month, and then deliver more in each month that follows.
Interestingly, seeing the software -- incomplete though it may be -- often
makes everybody realize that perfection looks different than they originally
thought.

If Tunes had a customer, and she demanded to see the results of the last
seven years' investment, what would you show her?  What if she had to show
one concrete feature to investors in three months... what would you build?

Jim