[microformats-discuss] Re: [Geowanking] geo microformat BOF session at Where 2.0

[Kevin Marks]

On Jul 6, 2005, at 1:57 AM, Edward Mac Gillavry wrote:

> Hmmm... Let's get us back on track, shall we? This list is about 
> GEOwanking. But maybe being a cartographer I don't get the 
> importance/relevancy of the discussion so far. I agree, there is a 
> chance for yet another representation of location to emerge with the 
> geo microformat. Therefore I'm happy this thread receives some 
> attention as I think there are many ways in which we can contribute to 
> a geo microformat.
>
> However, before going into the implementation, e.g. 
> XML/RDF/SomethingElse, what have we come up with so far in our 
> previous discussions? Should the geo microformat focus on point 
> locations only? Because that's what seems to be the main approach: how 
> to structure an address and translate that into a point location? 
> However, I seem to remember people on this list have realised it's 
> useful to represent area features as well. While we're at it: 
> polylines? Should we give an indication of accuracy, spatial 
> precision?

A key use case for a geo microformat is for people to be able to refer 
to a place from a web page. Currently, people do this predominantly in 
two ways:

  * referring to a human-readable address ('San Francisco', '665 3rd 
Street, SF, 94107', '3rd and Townsend', 'by the ballpark'
  * linking to an existing map service with a proprietary URL (mapquest, 
google, yahoo)

These are useful in human terms, but they don't really map onto a point 
location, as they are referring to a human-scale area.
Bare latitude/longitude is insufficient to present a cartographic 
representation to a user, as there is no way to know what scale to 
display that map at.

What would be useful, in my view, would be to translate the above into 
a latitude, longitude, and radius of interest. These can be translated 
back into different map display systems.

Translating the human-readable names is a tricky problem, but one that 
mapping services already tackle with variable success, so  having a way 
of presenting the human-readable and lat/long/radius data as 
alternatives is valuable.