[MG] Visualization problems for crossforum theatre

[Marcos]

On Sat, Jan 1, 2011 at 11:04 AM, Michael Allan <mike at zelea.com> wrote:
> I'm running into obstacles here:
> http://zelea.com/project/votorola/a/crossforum/axial/TriaxialPollMapV.svg
> That's a map of polls (issues) on three axes: activity (x), consensus
> (y) and turnout (z). ?A partial implementation is currently running:
> http://u.zelea.com:8080/v/crossforum/#c=DumT
> http://u.zelea.com:8080/v/crossforum/#c=DumT&s=5678
> But I can't finish coding it, because of visualization problems:
>
> ?1. It simulates the z-axis (turnout) by plotting circles of varying
> ? ?density. ?But that isn't convincing enough. ?We need a better
> ? ?visual for the z-axis.

Use a fog and scaling algorithm with your black background -- together
they should be sufficient to connote depth.  See the processing.js
visual programming language.


> ?2. It uses decimal offsets to show zoom level and pan position. ?This
> ? ?saves space, but it limits the zoom controls to factors of 10
> ? ?(10x, 100x, etc). ?That's probably too coarse. ?But if we allow
> ? ?finer zooms, then how do we visualize the zoom level and pan
> ? ?position?

Use a single motion-parallax bumb algorithm (not unlike nintendo
kinect) that will give the brain a depth and scale vector in which to
visualize and interpolate the rest.

> ?3. It's not immediately obvious what the map is showing. ?Maybe we
> ? ?can make it interesting/attractive enough that the user is willing
> ? ?to explore a little further, and eventually discover what's being
> ? ?shown. ?But how exactly?

These techniques are exactly what pangaia.sourceforge.net is going to
use for its UI, I hope we can work together sometime....

> ?4. How would you go about visualizing the landscape of democracy?
> ? ?What's the most likely visual to succeed at the overview level?
> ? ?If you can draw it, we can probably code it.

Don't look for (*the*) "one way", there are many perspective views
into the same data, each will communicate very different things
depending on the intent of the user.  Here's one main "principal
component" that is sure to emerge:  Project-centric view vs.
People-centric view.  There's no way to combine these in a single
"view" because they are orthogonal to each other in a way that cannot
be visualized at once without collapsing it all into a big huge,
unparsible mess.  Consider it like the need to keep yin seperate from
yang.

Would love to dialog more about what you're doing....

marcos



Originally posted to the mailing list of the Metagovernment Project:
http://metagovernment.org/mailman/listinfo/start_metagovernment.org