Re: Action 10 - What Is A POI? - "Place-Oriented Information"?

Hi Henning,

thanks...that helps me understand where you're coming from 8)

So are you suggesting that POIs be bound by the same namespace
requirements for civic addresses as outlined in these types of
rfc's[1][2][3]?

I can definitely see how some network contexts require an explicit and
normalised namespace...but I'm not sure that POIs are such a context.
Surely they would be free to utilise those references...but not
necessarily be the key to enforcing their uniqueness.  e.g. wouldn't a
POI just refer to a PIDF-LO document, WOEID[4] or similar?

Have you got some specific Use Cases where you think this is critical?

I think it was Dirk at the Mobile AR Summit that pointed to Tourism as
the archetypal Mobile AR Use Case and I believe that shows unique
addresses shouldn't be "mandatory".  e.g. "Captain Cook's first landing
site in Australia"[5] may be a significant place for Tourists but does
not really have a clear "civic address".

#poiwg NOTE: Unfortunately Tourism isn't really listed in the Use Cases.
"3. Take me on a scenic wine tour"[6] is probably the closest.

Henning, perhaps you'd like to clarify how you think PIDF-LO
descriptions relate to point 5 of the Data Model page[7].


roBman


[1] https://datatracker.ietf.org/doc/rfc4776/
[2] https://datatracker.ietf.org/doc/rfc5222/ 
[3] https://datatracker.ietf.org/doc/rfc5139
[4] http://developer.yahoo.com/geo/geoplanet/
[5] http://goaustralia.about.com/cs/nswsightseeing/a/botanybay.htm 
[6] http://www.w3.org/2010/POI/wiki/Use_Cases#Destination_Selection 
[7] http://www.w3.org/2010/POI/wiki/Data_Model#Data_model_for_a_POI 
 

On Sat, 2010-11-20 at 10:22 -0500, Henning Schulzrinne wrote:
> I've worked for a number of years in the IETF ECRIT and GEOPRIV working group, which have defined a range of location-based protocols. Current use is primarily in emergency communication (NG911). See RFC 5222 et al. Thus, my "bias" tends to be large-scale, long-lived systems that need to interoperate reliably.
> 
> Henning
> 
> On Nov 19, 2010, at 11:45 PM, Rob Manson wrote:
> 
> > Hi Henning,
> > 
> > sounds like we definitely have a context mismatch 8(
> > 
> > I think Gary's summary[1] (which was really the basis for this whole
> > thread and my response to Dan) laid out the distinctions quite well.
> > Although as "Director, Ovi Places Registry" he's obviously coming from a
> > loaded context 8)
> > 
> > Out of interest...what kind of cultural context are you grounded in?
> > 
> > My background is Augmented Reality and Pervasive Computing which has
> > shaped my views on Space, Place and Location.
> > 
> > 
> > roBman
> > 
> > [1] http://lists.w3.org/Archives/Public/public-poiwg/2010Nov/0044.html
> > 
> 
> 
>

Received on Sunday, 21 November 2010 03:06:01 UTC