Names and Identifiers for Physical Locations [UPDATED]

Earlier this week I managed to miss a talk about Erewhon, an Oxford University project that “exists to research and develop methods of accessing University information and services from mobile devices”. The project maintains, amongst other things, oxpoints, a ‘name and location’ server that provides a dictionary of geocoded names relating to Oxford University sites and departments. (Fortunately, @ostephens blogged the talk here: Telstar blog: Erewhon).

It was interesting to see one of the early mashup examples on the Erewhon blog was a Simple Library mashup that allows you to “(f)ind the nearest copy of a book from a reading list” using the oxpoints service.

The whole ‘bringing places to the browser’ took another step forward today as well with the announcement of Place Pages for Google Maps. The idea appears to be to have a (wiki like?) page for every location… but what particularly caught my eye was the structure of the URIs, which are human readable and based on the actual address… So for example:

which you’d probably expect to be there… and locations like:

which you possibly wouldn’t?

So now you can guess at the human readable, and human hackable, URL of a particular location.

PS each location also seems to have a machine URI, rather than the human readable URIs, containing a cid argument which I’m guessing is some sort of place identifier? For example:

It probably bears no relation to Yahoo’s WOEIDs, though, which are also used by flickr, but it would make sense for someone to start publishing relations between them?


Hmmm – that Downing Street link just goes to a placeholder page…

But this more detailed page also exists:

That is, after the address give the ‘business name’, in this case:

So if you have an address – e.g.
and you add “-BUSINESSNAME”, you get a business page?


What seems to be happening is that the first part of the URI provides a geographical context for a business search term.

So for example:

That is, the URI seems to be implementing a 2D search: a location part (eg gb/london/mayfair) and a business name part (harrods, identified by the leading )?

(For more 2D search thoughts, see: Mashlib Pipes Tutorial: 2D Journal Search or searchfeedr.)


  1. Pingback: Google introducerar place pages :
  2. Owen Stephens

    I don’t know if you saw my follow up to the Erewhon talk on the Telstar blog ( I note a couple of services that might help with identifying library locations – although as far as I know neither are complete. I also note the issue of dealing with electronic access to resources and the idea that a user actually works in several contexts simultaneously, and usually no single institution knows about all the contexts.