Thanks to Martin Stabe for linking to the news from Steve Outing that Adrian Holovaty will release the code behind Everyblock.com as open source. Everyblock is the Knight-funded project that Adrian launched last year. It’s an extension of his groundbreaking ChicagoCrime.org. Normally, I would just link to this via Del.icio.us, but this development is important enough to warrant a post. It’s Christmas in June, when Adrian said he will release the code as open source. This is a huge gift from Adrian and by extension the Knight Foundation, if only the industry will take it. Steve Outing has highlighted one application for Everyblock: Mapping classifieds.
When asked about the utility of mapping classifieds, Holovaty says that absolutely that’s a great use for Everyblock’s system. You can easily imagine the Everyblock concept applied to garage sale ads, lost-and-found listings, or real estate listings, for example. Mapping could be useful for other classified categories, too, but of course there are privacy issues to consider. (A private classified advertiser selling his car or bicycle might not want his ad showing where he lives, but since there’s a benefit to exposing that information, he should be given the option.)
There are other valuable points in Steve’s post. He is not alone in encouraging news organisations to begin thinking about geo-tagging. It’s not even that difficult to add geo-data to stories semi-automatically by parsing the data from the dateline, although the more local the story, the richer the data one would want to add. An increasing number of mobile phones include GPS functionality, and it’s not difficult to add this to story data. What about a script that would automatically add that data if the journalist used the data connection via bluetooth to file? We all should be considering adding a geo-data field to our databases. It’s a basic step that enables a vast range of journalistic and commercial applications.
And Steve and Adrian go one step further saying that geo-tagging is only one bit of meta-data that news organisations should be consider adding. One of the subtle changes the Guardian (my day job) has made to its CMS is to think about the information architecture of the site and the stories. Structured data, as Adrian has shown in several projects, allows news organisations to begin to make sense of vast amounts of information in novel and very useful ways. We can use our own meta-data to help show trends, make connections and add context to our journalism.
As Steve says, there are alternatives to Everyblock, and the code may not be applicable to all projects. However, I’m increasingly concerned about a ‘not made here’ complex amongst journalism organisations. You can see some of that in the comments on Ryan Sholin’s recent post about building a local news site from scratch. Some of the comments are critical of open-source projects Drupal and WordPress. I’m not a knee-jerk open-source advocate, and some of the criticisms are no doubt valid. But I think many open-source projects deserve consideration along with the custom CMS route. Morris Digital Works has done some ground-breaking work with Drupal. The New York Times has even gone as far as being a good open-source citizen by releasing some of their code on their Open blog. All the code that’s fit to printf(), as they say.
But I’m convinced that news organisations have more to gain than to fear from open source and projects like Everyblock. Open source can be another form of networked journalism. Instead of relying solely on your own development team, you can suddenly plug into a worldwide network of passionate developers.