Tim O’Reilly, Web 2.0 and Geospatial

I was probably the most excited member of Directions’ staff when I heard Tim O’Reilly would be giving the keynote at our conference. I’d watched my programming colleagues (yes, Peter, that’s you) use the awk and sed books early in my career and more recently began reading the many O’Reilly blogs online.

While O’Reilly covered a number of big ideas (which I hope to cover elsewhere in Directions) I want to highlight here some of the “rules for success” of what O’Reilly calls Web 2.0, the next smarter, more interactive, more useful version of the Web. In particular I want to look at the rules in terms of geospatial which highlight some of what geospatial is “missing” perhaps.

1) Users add value - Think of how Amazon taps into not just what users buy (“this book is ranked 149,000”) but their opinions on the book via reviews. Many, many of us use Amazon’s features for purposes other than buying books (and guess what? That’s ok by Amazon!). Can and do users add to mapping websites? Just barely - new model sites like Flickr and Mappr and A9’s mapping tools do provide ways to contribute, but most sites are strictly presentational: “here’s a map; have a nice day.”

2) The network effect should be a default - Website developers should take value from web visitors, even if they choose not to provide a review or check a box about how much they liked an article. Napster had it right, O’Reilly noted, since once you downloaded a song, you contributed by now being “required” to serve that song. I suspect there’s some geo market research data being collected this way (where you are from based on your IP address) but it’s not serving the “greater good” as the Amazon “most popular” or old Napster did.

3) Experiment - There is a reason so many of Google’s services are Beta - they are experiments. And, thus they can come and go or mature slowly, while still be in cutting edge. I think we in geospatial get this one, though I have yet to see a city mapping website that says “beta.”

4) Software above the level of a single device - These days its not enough to develop a software product for a single platform, it needs to be ready and able to move to other platforms and form factors. Can we take maps with us from the Web on handhelds? Only sort of, I’d argue.

5) Data is the next “Intel inside” - So, from a business standpoint if you want to be in the technology stack for the long run either buy or create a hard to recreate dataset (think NAVTEQ, Tele Atlas, for example) or be the domain owner/organizer (“own the namespace”) for that data (like Westlaw for legal docs, or CDDB the online database of CD and songs). Can you think of a company trying to do that in the geospace? How about ESRI’s Geography Network? Or, perhaps its sister ESRI implementation, GOS?

6) Platform will beat an Application Everytime - Excel and Word didn’t win; Windows did. I’d even offer that vertical applications on AutoCAD didn’t win, but AutoCAD did. Ever wonder why so many technologies today are not products but platforms? Platforms are the underlying soil for applications. Google is a platform, no? Google Map is a platform, too, one rife with excited hackers doing neat stuff.