[FoRK] [Tech] Goggles, or Projectors?

[J. Andrew Rogers]

On Jan 2, 2008, at 5:40 PM, Lion Kimbro wrote:
>  * protocol -- we need regular / standard protocols for figuring out
>                where things are, what they project/see, how things
>                are moving, etc., etc.,.


Easy in theory, hard in practice.  On one hand you have the problem of  
measuring reality as it unfolds and representing that on a computer  
with high fidelity (it sounds easy in theory, but it does not scale).   
On the other hand, no one can decide on useful protocols and  
standards, scaling issues aside.  A lot of people are working on what  
is essentially the exact same problem in a diverse range of spaces,  
but they are generally not working together and are optimizing for  
particular domains.


>  * bandwidth -- there needs to be sufficient bandwidth between
>                 components to maintain a steady-enough persistent
>                 illusion.


You only need enough bandwidth to push video frames at some arbitrary  
frame rate.  There is an argument to be made that at high enough  
detail, it is cheaper to render on the back-end.


>  Just about all of this exists *somewhere,* but it's not integrated.
>  Someone wrote a paper over there, Myron Kroeger had made something
>  else over there, so-and-so has an interesting system over there.


Again, not as easy as it sounds.  The US DoD has perpetual RFPs open  
for this type of technology because they hit the scaling wall over a  
decade ago. The limitation is not hardware or interest but algorithms.


>  I think it's all **opportunities**, until one of us makes it
>  happen!
>
>  Fortunately, I don't think it's very hard to get started;  It just
>  requires time;  Time, and faith.


Riiiight...  Computer science may take time, but all the faith in the  
world won't alter the computational complexity of an algorithm.

Cheers,

J. Andrew Rogers