[Sigia-l] Mapping

[karl fast]

> > > Another "visualization" tool to put on the scrap heap next to
> > > antarcti.ca, ThemeScape, kartoo, and lord knows what else.
> >
> > You're suggesting that visualization is a bad idea, and I think that
> > is a bad idea.
> 
> Sigh. Well, you can extrapolate that.

Ahh, well I made a statement and should have asked you a question:

    Are you suggesting that visualization is a bad idea?

That's how I read it and I don't agree that visualization is all
bad. Reading your other comments I think we're on the same
wavelength (I'll be damned!).


> I guess I'm getting annoyed with how the shiny-designy solutions,
> which aren't useful, are getting noticed, while the less pretty
> solutions, though actually useful, are by-and-large ignored.

This relates to my concerns with information visualization.

Near the center of info-viz is the notion that as the amount of
information in the world increases we must move away from classified
lists of text in order to navigate it. Classified lists of text are
low-bandwidth. The human visual system is high-bandwidth and can be
used to exploit our extraordinary perceptual abilities.

This is all good in theory, and even in practice. But the evidence
is that classified lists of text are doing just fine, even for large
amounts of information. Consider Yahoo, Google, Ebay, and Amazon.

I would issue three caveats to this:

 1. I make a distinction between tools and techniques. Things like
    Groxis are visualization tools, but many of these techniques can
    be applied in systems where they form a key element of the
    solution instead of being the central paradigm.

 2. Info-viz has promise, but I'm not sure we've figured out enough
    to develop mass-market visualization tools. And we may not, but
    this is an area where basic research needs to continue.

 3. The private sector always wants to exploit this basic research.
    To do this they must guess when something is ready to make the
    leap from the lab to product. Some people have to lead here and
    eventually it someone will figure it out. If Groxis works,
    great. If not, I just figure it's part of the
    research-implementation-producization churn.

    

--karl