[LINK] limits of technology in finding someone

[markus.buchhorn at anu.edu.au]

Roger wrote:
> Back to the point:  in classical Information Systems theory terms, 
> GPS is a decision support tool that's being used by naive people as a 
> decision tool.

Right, but you need to be a bit more careful with the language here. 
There are actually three components here: the GPS satellites, which 
provide XYZ coordinates to the best of their ability, the mapping data, 
which can be a bitmap scan for positioning information or vector data for 
routing information, and the software that combines the two and provides 
the user with some information that they can then base a decision on if 
they choose to.

I've used GPS on my mobile phone (o2 xda) with a bluetooth gps receiver 
across Australia and Europe on well over 100 drives, probably a few 
thousand km, so far only with vector-based mapping data for routing 
(driving) instructions. I have had exactly one occassion where the GPS 
receiver let me down (I think it got overheated in the Sun; powercycling 
fixed it). I have had probably two dozen momentary errors due to dodgy 
mapping data (laneways treated as roads, new roads being opened/closed - 
and I can buy new data every 6 months or so), and probably less than a 
dozen where the software has decided something weird and I can't be sure 
it is bad mapping data (where turnoffs are actually merges, or road-name 
changes). In most cases I have spotted the problem up front and ignored 
it, or just missed a turn and recovered a bit further down the path.

I haven't used it on bitmap mapping (scanned maps), but the reported 
coordinates and paths are reliably spot on within the accuracy of the 
receiver.

So when we talk of 'gps failures', bear in mind there are several bits 
involved. Vendors don't make that clear either <sigh>.

Cheers,
      Markus