[FoRK] map search
J. Andrew Rogers
andrew at jarbox.org
Sun Mar 3 19:31:07 PST 2013
On Mar 3, 2013, at 1:40 PM, Gregory Alan Bolcer <greg at bolcer.org> wrote:
> Good point. Actualy Layar is like an onion--lots of layers which pretty much screws up any usefulness.
The idea of layers is a useful organizing principle and easy to conceptualize in a spatial context.
What all of these platforms lack is the ability to dynamically connect layers with operators to produce a derivative "view" layer, at least on any scale that matters. A big hurdle is that most of the algorithms and databases used to build these platforms aren't capable of usefully delivering this functionality.
> Google Indoor Maps are very promising, though they need to support and indoor augmented reality model rather than the spatial translation between flat and real.
> Speaking of which, the ability to read flat maps and translate what you are seeing into realspace/3d is often cited as a top issue. In fact, I think there's a bunch of meme gifs making fun of people that I'm too lazy to look up.
Building a general purpose 3-dimensional geospatial model at scale with good registration would be very cool but is technically beyond the scope of what anyone is currently building. There are a number of difficult computer science and software design problems that show up when going from 2-space to a 3-space that do not have good solutions. Very few organizations can do 2- or 2.5-dimensional well as it is.
> We used the Justin Bieber example too, not because it was common, but because he really did show up at Pizza Port at that exact bench.
Justin Bieber is the "Hello World" of social media search, much like how Starbucks is for points of interest and Doppler radar is for sensing data. For better or worse, Justin Bieber reliably has a relatively broad and dense distribution in many real-world data sources that make it good for testing and demos.
More information about the FoRK