[georss] Geolocation by reference

[Sean Gillies]

Carl,

Thanks for the pointer. Good stuff going on there, but I don't see that
any of it is immediately useful to the Atom/Geo community (or my little
classicist corner of it).

Sean

Carl Reed OGC Account wrote:
> Sean -
> 
> There is a rather large knowledge base on location by reference and
> standard approach that have been discussed and well defined by the
> GeoPRIV WG of the IETF (www.ietf.org/html.charters/geopriv-charter.html
> ). This work is a component of the definition of a standard expression
> of location objects (LO) for use in a number of internet standards.
> There are a number of internet standards (HELD, SIP, RADIUS, PIDF etc)
> that already reference as mandatory the formal expression of the LO. The
> expression may be by civic location (such as an address), by geodetic
> location (coordinate geometries using a GML application schema) or
> location by reference (which can be a combination of a number of ways of
> expressing location). These standards are being (or have been)
> implemented by the internet and communications infrastructure companies.
> These standards will also be an integral component of the NG 9-1-1
> implementation.
> 
> Some of this work may be of help in what you are trying to do. Also,
> would be nice to stay aligned with the internet community :-)
> 
> Regards
> 
> Carl
> 
> ----- Original Message ----- From: "Sean Gillies" <sgillies at frii.com>
> To: <georss at lists.eogeo.org>
> Sent: Friday, November 09, 2007 4:59 PM
> Subject: [georss] Geolocation by reference
> 
> 
>> GeoRSS has literal locations well covered. A possible use case for
>> location by reference (hyperlink really) is coming up in my work and I
>> want to run it by this group.
>>
>> My Pleiades project is collaborating with other digital classics
>> projects to build out an ancient history web. Pleiades aims to be the
>> authoritative gazetteer for the Greek and Roman civilizations; we
>> maintain place name and locations resources that can be linked to from
>> other projects for geographic context. For example, see this page on the
>> American Numismatic Society web site about a coin from the Xanthos mint:
>>
>> http://publicserver.numismatics.org/collection/accnum/list?accnum=1977.158.477&single=1
>>
>>
>> The developers of the site are harvesting coordinates from the Pleiades
>> Xanthos record (http://pleiades.stoa.org/places/639166) and will be
>> turning them around for use in their own GeoRSS feed (of mints). It
>> works, but it's a fair amount of effort that has to be reimplemented
>> from site to site across our little history web.
>>
>> I think there may be cause here to add more declarative syntax to the
>> georss:where element. My initial idea is that georss:where could have an
>> optional src attribute exactly as atom:content can. The value of the src
>> attribute should be the URI of a GML document. Like so:
>>
>>  <atom:entry>
>>  ...
>>  <where src="http://example.com/locations/1.gml"/>
>>  </atom:entry>
>>
>> where the resource at http://example.com/locations/1.gml would be
>>
>>  ...
>>  <gml:Point>...</gml:Point>
>>
>> In this way the numismatists can reuse the authoritative Pleiades
>> locations and need neither maintain their own duplicate database nor
>> continually synch against Pleiades resources. The synchronization could
>> become built in.
>>
>> There is precedent in Atom for remote sourcing, but only to my knowledge
>> for atom:content. The src attribute feels pretty good to me, though I do
>> think that it has the potential to make location (which is metadata) a
>> bit too much of content (data). Does anybody have other ideas for
>> non-literal locations?
>>
>> Sean
>>
>> _______________________________________________
>> georss mailing list
>> georss at lists.eogeo.org
>> http://lists.eogeo.org/mailman/listinfo/georss 
> 
>