Vilfredo celebrates Christmass with a big update

[Pietro Speroni aka Vilfredo]

With the Christmass vacations we had the time to really make some big
progress on Vilfredo.

While Derek focused on Captcha, and security issues, I mainly worked
on introducing Graphviz images inside Vilfredo.

I always wanted to do it. It seemed Vilfredo voting structure was just
begging for it. And indeed once we made them, the result came out
awesome. Still a lot of work needed to be done to create images of
different size, introducing all sort of simplifications, to make the
graph "as simple as possible, but not simpler" as Einstein would have
said. Each voter in a generation of a question was represented as a
darkpink egg, while every proposal with a box. A blue box if the
proposal was in the pareto front. The position of the box would tell
us how many people voted for that proposal, and a black arrow would
indicate which proposal dominate (if above) or were dominated (if
below) by it. After a bit of training (I intend to make a screencast)
a person looking at one of those graphs can see who voted for what,
what are the pareto fronts (that is immediate), which proposals have
been voted by the same people, and which people only voted for a
single proposal (thus forcing it to be in the Pareto Front).

A lot of work went into making the image readable, and finding a good
compromise between having all the information I wanted to pour in it,
and keeping the image not too complex. And if Vilfredo starts to be
more popular, with more people using it, those images might still
become messy and spaghetti-monster like.

We then introduced in the images the possibility to highlight
particular proposals (thus highlighting all the users that voted for
that proposal as well as all the proposals above and below), and the
possibility to highlight particular users.

One fo the interesting side effect is that Graphviz dot program
started to place near proposals that are voted by "about" the same
people. Which often have also a lot in common. In one case I spotted 3
proposal that were bundled together by graphviz and that were
basically saying the same thing in different ways!

I think we are really onto something with those images, and the
intuitive understanding of the space of the proposals that the image
give. It's still a 2 dimensional structure. It probably wouldn't work
with too many proposals, some times it works better, soime times it
works less well, but all in all it is a great progress.

Then we integrated those images in the pages of the site.

The proposal page was fully redesigned, to show all the generation a
proposal went through. (A winning proposal goes from one generation to
the next).

A generation page was introduced with all the information about that
particular generation. This page was particularly challenging, as I
managed to fit into it also some information about "alternative
history". In other words, if this person did not vote, how would
things be different? Was there a way for this person to make the
Pareto Front smaller?

In a normal voting democracy, no person alone has real impact on the
voting time. This is not necessarily true when the process uses the
Pareto Front. So we go and look for each person in what way (if any)
their presence changed the resulting Pareto Front.

It's all very new,
it's all very exciting,
and it all needs a lot of testing.

Oh, yes, we also changed the history page.

You can see an example of this work looking at the pages:

http://vilfredo.org/viewquestion.php?q=67&room=Vilfredo
http://vilfredo.org/viewhistoryofquestion.php?q=67&room=Vilfredo
and for a particularly interesting generation look at this one:
http://vilfredo.org/viewgeneration.php?q=67&g=1&room=Vilfredo

But we haven't finished yet.

Cheers,
Pietro