Breaking down the barriers to mapping
It seems like the mapping and GIS world is opening up to a much larger set of possibilities.
I've spent the last couple of days at the Location Technology and Business Intelligence conference, listening to people talk about all kinds of location-related computing, mostly aimed at the enterprise. I asked a question about fun, and people seemed to respond pretty well to that - things like the Google Maps/Craigslist combination have this group of large scale integrators, data providers, and data managers thinking once again that mapping has a tremendous cool factor.
Google's done something remarkable in building a freely accessible and somewhat extensible mapping application on high-quality data. Most people use maps, on the web and on paper, but it's generally been easier to mark up a paper map than an electronic map. (I've seen lots of paper printouts of MapQuest maps with pen annotations adding critical pieces.) Google Maps makes that kind of work vastly easier, though no doubt I'll be seeing marked-up Google Maps printouts for a few years to come.
Google Maps feels to me like it's accelerating a process that had already begun. Google's deep pockets for data and computational power, not to mention its tremendous reach, made it easier for Google to produce something both powerful and fun than anyone else working in the field. Google's implementation choices have made it possible for people to create all kinds of things, like the Google Maps-showing-craigslist-rental-properties combination, without needing to know all of the usual details about projections, datums, and data quality.
How long this will last is anyone's guess - all that high quality map data can't be cheap, and it's not clear what Google's trying to accomplish here. Even if it doesn't last, however, it's shown a lot of people what mapping should be like - easy, interactive, and markable. Letting people put their own pins on a map isn't a grand technical innovation, but it solves a huge number of problems.
If Google Maps has you excited, there's a lot more to come. The price of mapping applications is plunging. That's partly thanks to open source applications like MapServer, PostGIS, GRASS, GeoServer, gpsd, OpenEV, and many more. Commercial vendors are also offering deals, like Manifold's $245 desktop mapping and web sharing application, Microsoft's $299 MapPoint 2004, and even Oracle's inclusion of spatial tools in most of its product line. Web-based services are also making it easier to do mapping on a transaction-by-transaction basis.
The barriers are coming down, though cartography, geographic data, and even these lovely tools aren't that easy to explain. (We're working on that at O'Reilly, with Mapping Hacks and Web Mapping Illustrated coming next month.) Figuring that out enough to create basic usable maps isn't so difficult, but there's another challenge ahead for would-be mapmakers: data.
Americans are relatively lucky, in that we have lots of data available from the federal (and sometimes state and local) government for free. That data isn't always of the quality we'd like, which ensures a market for Navteq and similar companies to provide their own more highly polished data at a price, but there's enough there to get started. If you build on the Census Bureau's TIGER data, you won't have the detailed overpasses shown by Google, but you'll still have a workable start. (Unless, of course, your roads were all built very recently.)
Not everyone is this lucky, of course, as many countries don't offer basic data like the U.S. does. Even when it's available, the free data usually has limited quality, but it removes one huge barrier to getting started. It seems like a perfect way to bring more people into mapping - reduce the initial costs of the data along with the tools, and give people a foundation they can build on. Users can then either improve the data themselves (more and more plausible, thanks to GPS) or buy improved data from the many vendors out there.
A lot of the parts for making mapping ubiquitous and easy, at least at the start of the learning curve, are out there. Now we just need to explain it, and convince many recalcitrant governments that holding on to basic data suffocates a lot of exciting and useful possibilities. Neither of those will be an easy task, but enough barriers have come down over the last few years that it seems to reasonable to hope a few more will fall.
Simon St. Laurent is an associate book editor at O'Reilly Media, Inc..
Comments on this weblog
Return to weblogs.oreilly.com.