Got a cool note from Jay Nath yesterday--he's the open-data pioneer behind San Francisco's DataSF.org--saying he made it mandatory for city departments to include metadata with any new datasets they upload to DataSF.

That's really great to hear, esp. since this development grew out of a discussion at CityCamp SF this weekend that I led with Tina Lee, a recent fellow at city's Department of Technology Services, and Mat Dryhurst of the Craigslist Foundation (see last week's post for background). About 25 people participated, including gov workers of various stripes, nonprofit folk, developers, and a few journos. Wanted to briefly recap the discussion for more people interested in how open-gov data is released and used.

Lot of talk about limits of capacity and awareness among city staff--relatively few people in city government know about DataSF yet (it's fairly new), and there's concern about "scaring off" new users by adding requirements. In the contest of quantity vs. quality (i.e. a DataSF with tons of hard-to-understand datasets, vs. just a few really well documented ones), Nath and others in attendence agreed that this growing young platform is most interested in maximizing quantity right now, which makes sense.

Couple good responses bubbled up: start a brown-bag lunch series where city staff can meet the citizens, developers, and journalists interested in city data, and provide an open wiki inviting citizens and city workers to ask and answer questions about the data.

I especially like the brown-bag suggestion, because it gets at the heart of what makes DataSF and other open-gov data initiatives so powerful for both the data owners and the data users out there. By releasing their data in useable, welcoming formats and actively inviting citizens to build upon them, city agencies can actually get closer to fulfilling their own core missions (see EcoFinder for a good example). Andy Maimoni of Open311 made this point really well on Saturday, and I think it's a selling point for why data dictionaries matter. Looks like the brown-bag idea has legs; will update here.

Side note: As a journalist, I found myself navigating a solid line during Saturday's discussion. While I'm an advocate of open-gov data, I'm also deeply interested in finding and visualizing the data that government really, really doesn't want to make public. I'm excited to see DataSF and other open-gov platforms evolve and grow, largely because I want to see what citizens build as more data becomes easily accessible and understandable. But open-gov platforms will always be just one of many ways newsrooms like ours get at information, along with FOIAs, public records requests, and simply talking to enough people, about enough stuff, for enough time.

Here's Tina Lee summing up our discussion, and thanks to all who participated: