tommy's scratchpad

Something’s gotta give.

There has been so much discussion on the idea of geospatial catalogs/registries/repositories/searching/[insert_buzzword_here] in the last little while that it’s impossible not to see the sore thumb that it has become within the community.

Discussions: Paul Ramsey has offered his experiences on this on his blog in October 2005, then again in February 2006. The Digital Earth Weblog also discussed this recently. I won’t get into the specifics of the discussions except to say that everyone agrees we need a simple catalog API/solution for the community. A lack of solid implementations and discussion about catalog profiles has gotten us to where we are at this point.

Implementations: The Geospatial One Stop (GOS) portal provides a CS-W currently. So does a prototype from Agriculture and Agri-Food Canada. Refractions has done some brave work in scraping Google to populate a catalog they implemented for uDig. owscat has been deployed in the ResEau portal as well as a UN Water prototype. GeoNetwork Opensource and deegree2 both promise CS-W implementations. And this is not to mention the barrage of ad hoc APIs out there already attempting to do the same thing.

Now, OSGEO is picking up some steam on this trail. The Geodata group is looking at requirements in defining metadata and a catalog (see their wiki). Many approaches are being reviewed, such as CS-W, pure OGC:WFS, and OAI-PMH, which is the Open Archives Initiative Protocol for Metadata Harvesting 2.0. This one is new for me, but certainly worth a look.

Hopefully these activities can somehow be unified, at least at the API level, for interoperability and federated peering. In my mind, the main goal would a simple catalog, with a “publish-once” approach for data and service providers, and the ability for lightweight clients to implement against. Seems to me like this would be a viable resource in the GeoWeb 2.0 paradigm, eh?