First results of the OGC - W3C “Spatial Data on the Web” collaboration
Despite all the temptations of summer, a recently formed OGC - W3C collaboration, the Spatial Data on the Web Working Group, has progressed to releasing the first results of its work. The Spatial Data on the Web Use Cases and Requirements (UCR) Document is being released jointly as a W3C First Public Working Draft and as an OGC Public Discussion Paper. The documented use cases are an effort to put a human face on where and for whom the group’s work should be useful. Some of the use cases include: Meteorological Data Rescue, Harvesting of Local Search Content, Locating a Thing, and Linked Data for Tax Assessment - a very diverse collection. The derived requirements will serve to guide the subsequent standards work and put some reasonable bounds on the scope of that work.
The document also represents the first few months of cultural and technical “acclimatization” within a group that has dual standing as both a W3C and an OGC unit. Many of the group's members have substantial experience in both standards organizations, most have worked with spatial data, and all are familiar with Web technologies. Even so, there are distinct differences in how the two organizations work, whom they see as their constituents, and what their priorities are in bringing forward the Web of Data. From getting familiar with Zakim (the W3C IRC "bot") to bridging some of the documentary differences (Word-centric vs HTML-all-the-time), we have had some interesting learning experiences.
Even more than the cultural differences, though, the diversity of technical outlooks has been a challenge for the group to work through in its mainly remote (email, chat, and teleconference) interactions. Everyone lives in space and time, but not everyone has concerned themselves with the geodetic, geometric and semantic details of how to design high-quality data to represent features of space and time. Do we make sure that every human and machine operating on the Web has access to these details, or do we hide the details and try to make spatial data as much like any other data as possible?
The group has at least come together in the view that there are few easy answers but lots of interesting questions. Our challenge is how to make spatial data as ubiquitous and natural on the Web as possible without putting too many people in danger in those cases where perception created by data clashes with the hard reality of the world we move about in. As a group participant, I'm excited to see what a wider audience will make of the ground rules we've developed in the UCR document. I'm also excited to move forward from this point toward some spatial data practices that make spatial only as "special" on the Web as needed and not more so.