[geeks] 3D desktop interfaces...

[Charles Shannon Hendrix]

Sun, 25 Jun 2006 @ 15:40 -0400, Nate said:

> What do you folks think of BumpTop and lowfat?
> 
> http://honeybrown.ca/Pubs/BumpTop.html
> 
> http://macslow.thepimp.net/?page_id=18

Interesting.

However, I never have thought that absolute desktop models were that
great.  Not everything works best when done the same way we do it in the
real world, visually I mean.  Nothing wrong with files and desktops, but
when you start basing your *actions* on a real desktop, the abstraction
begins to leak pretty quickly.

One thing I noticed in the demos is that while it looked really neat, I
was never able at any point to tell *what* data was being manipulated,
nor was I able to tell what relationships there were between the piles,
except for type.

In other words, I saw nothing I cannot already do with a normal file
browser, except have it look cool and eat tons of CPU/GPU resources in
the process.

I also saw a lot of features that were missing.

People have, for about 30 years now, talked about how we *must* replace
hierarchical storage.

I have three primary comments:

1) Most computer filesystems and desktops let you do a lot more than just
hierarchical, and in fact most of what people call that, really isn't.

2) We have been filing and classifying documents pretty much the same
way for 3000 years (that we know of), and the reason for the lack of
change is not because we've not tried, but that most other systems
ultimately failed.

3) In both demos, they were working with a relatively small number of
files.  I could see where working with a lot would require a good
mapping/panning system, and until they also add that, I don't see this
thing scaling beyond a single folder of data.

To me, the primary benefit of further research is going to be found in
metadata storage, not changes in data storage.  We can already do a lot
better than we currently do, but there are no standards in place that
can be used everywhere.  Also, I think metadata will be in layers, and
that we will be best served by keeping the "file" and "directory"
metaphors, but also add other methods on top of that.  There are just
times when the simplest of views works best.

If the WWW had not been jumped on so quickly, new generations of
hypertext metadata visualization which were years ahead of HTML and the
WWW could have caught on.  Instead the continuing fad of the WWW has all
but eliminated further research into that area.  In particular, Veronica
2 and an Andrew System data browsing facility, literally 20 years ahead
of the WWW, died almost overnight when the WWW caught on.

> (To me, they seem only marginally better than MS Bob.)

Hmmm... I saw no resemblence at all.

With some additional features, I could see it being useful for certain
kinds of projects, but as a universal tool, I don't think so.

> - Nate
> _______________________________________________
> GEEKS:  http://www.sunhelp.org/mailman/listinfo/geeks
> 

-- 
shannon "AT" widomaker.com -- ["The trade of governing has always been
monopolized by the most ignorant and the most rascally individuals of
mankind.  -- Thomas Paine"]