A few links that may be helpful:
It may be useful to be able to geolocate content in rss by brute force - beyond even well formatted microformat content - see:
http://ondemand.metacarta.com/
http://ondemand.metacarta.com/client_samples/geosearch.js.htmlMay be useful to visualize calendars on a timeline, not just a calendar view:
http://simile.mit.edu/timeline/Another tool to help with venue discovery is the new Zillow Neighborhood database. It has been hard to know the geographic polygonal boundaries of colloquial neighbourhoods - and such data may help; although of course this implies eventually having your own database of geographic data - which starts to leave behind off the shelf turnkey solutions like google.
I came across this today by chance:
http://spongecell.com/Beyond this, clearly the calendar database will want to build up an understanding of 'places' or 'venues'... One thought is that is it possible to perhaps define a UUID for places that extends beyond the application itself? Of course DNS defines a global worldwide unique identifier scheme for subjects - and it is probably best to use that... For example can Wikipedia descriptions or Placeopedia descriptions be used as the canonical identifier for a venue? Canonical identifiers help because if they are discovered in a feed then that automatically provides the venue location... It helps data escape the silo... I don't know if evdb or other event sites try to define a UUID.
- a