Re: AW: [Geopriv] Quickrandomcommentsondraft-ietf-geopriv-pdif-l o -profile-00

James M. Polk wrote:
> Henning
> 
> I wasn't claiming "mine is better", I was just defending 3825 as 
> suitable, when more than one seemed to claim it was not.

The other James was.

> 
> I agree we need consensus on what level of precision for each format 
> needs to be present to satisfy certain functions (like routing to a 
> PSAP). That seems like a meaningful discussion - and hopefully a more 
> peaceful one  ;-)

Since the PIDF-LO data may be used for functions other than PSAP 
routing, this might actually belong in an ECRIT document, but I don't 
much care either way.

For geo, the problem is that different jurisdictions have different 
requirements for accuracy, e.g., the FCC rules for wireless devices that 
speak of probabilities for different accuracy levels. Since 
jurisdictional boundaries for PSAPs are typically aligned on property 
lines, i.e., individual residences, for residential or commercial use, 
you could have a 1 foot error put you erroneously in the wrong county 
and PSAP. I don't think this matters in practice, but it indicates that 
picking a number will be rather difficult. I suspect that in almost all 
cases, people will simply provide whatever accuracy they can get from 
their GPS or other location device and set the bit count to "max", 
regardless of whether this is actually true or not. (Converting the 
circular accuracy estimates you get from your GPS unit to the units used 
in 3825 requires a bit of spherical geometry calculations that I don't 
see your average local wiremap maintainer performing.)

Henning

_______________________________________________
Geopriv mailing list
Geopriv@ietf.org
https://www1.ietf.org/mailman/listinfo/geopriv