Thursday, December 17, 2009

Interoperation with Pleiades

I've had a few questions lately about how other web-based publications could be designed to support interoperation with Pleiades. Here's my working advice:

Any project that wants to lay the groundwork for geographic interoperability on the basis of Pleiades should:

1. Capture and manage Pleiades identifiers (stable URLs like for each place one might want to cite.

2. Request membership in the Pleiades community and add/modify content therein as necessary in order to create new resources (and new URLs) for places that Pleiades doesn't yet document, but which are provably historical and relevant to content controlled by the external project.

3. Capture and manage stable URLs from Wikipedia or GeoNames that correspond to modern geographic entities that are relevant to the content controlled by the external project. Don't conflate modern and ancient locations, as this will eventually lead to heartbreak.

4. Emit paged web feeds in the Atom Syndication Format (RFC 4287) that also conform to the guidance documented (with in-the-wild, third-party examples) at:

and make use of the terms defined at

to indicate publicly relationships such as "findspot" and "original location" between the content controlled by the external project, Pleiades resources, Wikipedia resources, GeoNames resources and resources published by other third parties.

5. Alert us so we can include the entry-point URL for the feeds in the seeded search horizon list for the web crawler and search index service we are developing.

You can see how the Epigraphic Databank Heidelberg team has been thinking about how to accomplish this at: