MapBrief™

During the late 15th century heyday of Portuguese exploration, King John II forbade the open distribution of any map or navigational chart pertaining to New World discoveries under pain of death. Locked in a global land-grab race with neighbor Spain, cartographic intelligence was critical to expanding political power and exploiting the riches of the spice trade. While the link between this knowledge and economic advantage persists, in the last 500 years we have at least evolved to where transgressions aren’t enforced by the sword but rather the small-print legalese of the modern day end-user license agreement.

The small print was very much on my mind last week as I read “Empowering Citizen Cartographers”, a piece penned by World Bank official Caroline Antsey that appeared in the New York Times.  It begins as a paean to the wonders of crowd sourcing, especially in response to disasters such as the Haiti earthquake where Open Street Map shone as the de facto source of authoritative cartography. But then a new agreement between Google and the World Bank is described, whereby the latter actively promote and disseminate cartographic information from Google’s Map Maker platform. While Ms. Antsey indeed intended to praise Open Street Map, she seems singularly unaware that the actions of her organization may well bury Open Street Map in the developing world.

Because the license is clear: all of the data, all the fruits of the labor of those citizen cartographers, is the property of Google.  To be viewed through Google mapping interfaces with source data available under conditions specified by Google alone.  By contrast, Open Street Map data–yes, the raw data itself–is easily available to any and all, for purposes both non-profit and commercial.

Sure, nothing here explicitly prevents Open Street Map from continuing its work, but let’s get real: the deep pockets of Google paired with the imprimatur of the World Bank that effectively steers its partner governments, universities, and NGOs towards using the Map Maker platform may very well overwhelm Open Street Map’s more grass-roots efforts.  Google has shown an eager willingness to appropriate the tactics and rhetoric of community mapping, and of late, admitting to a bit of dirty pool in Africa against local startup Mocality.

(Google courageously pointed the finger not at its own employees but rather outside contractors it had hired.  Interestingly, if we were talking a violation of its own Map Maker terms, the old blame-the-contractor shtick wouldn’t play as this choice bit of Google language makes clear:  “If you are an entity, you acknowledge and agree that you are jointly and severally liable for the actions of your employees, contractors, agents, and other representatives. ” What’s good for the gander isn’t good for the goose, apparently.)

What’s in it for Google? Nothing more than a huge competitive advantage in the exploding smartphone market (and the concomitant local advertising revenue) in the developing world.  Imagine the commercial benefit of having exclusive access to the most detailed local cartography, collected for a pittance on the backs of “citizen cartographers”?  Even Tom Sawyer would blush.  It’s neocolonialism-meets-neogeography, only this time the shiny trinkets being dangled are laptops and Android phones.

If the World Bank was so impressed with the role of Open Street Map in Haiti, why throw its considerable weight behind the profit-seeking Google?  Who knows?  There have been collaborations in the past, and there appears to be a certain degree of chumminess in those circles. For those assuming Google is the only entity with the technical expertise to pull off the management of the crowd sourcing effort at this scale: please, stop.  Not only does Open Street Map have a platform and a track record, it also has the cooperative support of not-small-entities Mapquest and Microsoft.  So let’s put away the image of Google nobly shouldering a digital white man’s burden in bringing the developing world into the technically enlightened 21st century.

Make no mistake, Google has been the primary accelerant in the web mapping explosion of the last six years and they have spread the fruits of their innovation far and wide.  But the grating self-regard, borne of ideals that are never acknowledged to be driven by a motive so base as profit, has very much reached its sell-by date. It’s a 30,000 employee company hurtling towards middle-age whose growth has lately disappointed Wall Street: the potential profits in the fast-growing developing world figure largely in its future prospects.

~   ~   ~   ~   ~   ~   ~   ~

In 1494, Portugal and Spain signed the Treaty of Tordesillas, whereby under the auspices of the Pope Alexander VI, the New World was split between the two Catholic powers ad majorem Dei gloriam. One fears that the institutional favor the World Bank is granting to Map Maker will very much work to the greater glory of Google in the developing world, but at the expense of the full, free, and open access to the valuable information created by its own citizen cartographers.

—Brian Timoney

With the extra reflection that comes with any new year, I’ve been pondering a peculiarity of the presumably exciting geospatial industry: no one likes their business model.  Forget the giddy enthusiasm of 4-5 years ago, with the promised federal cutbacks at DoD/Homeland Security, along with the in-progress shrinking of state and local budgets, many shops are wondering how to keep treading water, let alone surf the wave of the next, ‘new’ thing.  How to explain this dissonance between a “cool” technology becoming more mainstream and the disquiet of not knowing how to profit from it?

Even though the Internet specializes in amplifying Moral Outrage, I’ve been taken aback by the public relations backlash against Google for having the temerity to charge its heaviest users of Maps. Well, more like reassured, since if Google (and Bing) has trouble explaining its pricing structure, then those of us who sell web-based services are allowed to cut ourselves some slack:

Reason #1:  No One Knows What Stuff Is Supposed to Cost on the Web

It’s been interesting to observe how the dominant vendor ESRI is playing their cloud-based offering. Since no one knows what things are supposed to cost on the web, using shrink-wrapped software analogs with which customers are already familiar helps…a lot. If you’re paying ‘x’ for an ArcServer license, then being able to do replicate the same end-user experiences using their online service for 60-70% of ‘x’ seems like a good deal.  A pronounced advantage to be sure, but competing on pure web experience is a punishing game, and so it’s even more important to lock in customers by any means necessary (including prodigious amounts of marketing).

*    *    *    *    *

There’s this upscale-ish farm-to-table place in my neighborhood where the wait-staff has been trained to regale first-time visitors with their ‘story’.  I’m hungry, I’m ready to drop coin, and your story is delaying my eating experience.  My enthusiasm has been converted to the singular wish that the wait-person just shut up right now.  I think we in technology are too often like that wait-staff, excited to overwhelm our customers with technical minutiae that fails to address their fundamental needs…

Reason #2: We Like Technology and Read Obscure Blogs; Our Customers Like Beaches, Kids’ Soccer Games, & Napping (and Don’t Read Obscure Blogs)

As an enthusiastic user/promoter of open source software, effusively digging into minutiae and wondering about business models is second nature.  Indeed, I recently received an email asking for advice on “open source business models” and my immediate thought was…I wish I had one. Luckily, someone much brighter than myself, Paul Ramsey, gave a great talk on this very topic at FOSS4G last year.  While Paul does a great job unpacking the complicated relationship between price and value, and how those signals can sometimes get very crossed, let me add a more general observation…

Reason #3: We Get Excited by Free, Cutting-Edge Technology; the Words “Free” and “Cutting-Edge” Make Middle Managers Very, Very Nervous

A couple of weeks back 60 Minutes profiled Alex Honnold, a guy who “free-climbs” cliff faces without any kind of safety equipment.  Brushing aside questions of safety and living with no margin for error, one was left aghast watching him calmly negotiate one life-threatening obstacle after another. Where he enthusiastically talked of future challenges, the viewer is unable to shake that this young man will meet a grisly, premature end.

Unfamiliar technology with unfamiliar licensing terms take many managers out of their comfort zone, without a safety net. Paradoxically, in a tough economy when their own positions are more tenuous, the appetite for anything resembling risk is minimal indeed.  I know, the availability of source code is the ultimate safety net. Have you seen the average manager’s pupils dilate in fear and confusion the first time you show them GitHub?

But let’s not make the managerial class the target of our animus, but rather evaluate honestly whether we’re opting to spend too much time in the World-As-We-Wish-It-To-Be instead of the World-As-It-Is.  Our clients and potential clients managed to stay in business before we showed up on their door step, so let’s temper the perma-sugar high of techno optimisim with a measure of old-fashioned humility.

In next week’s post I’ll discuss my best guess as to the key components of the still-elusive 21st Century Geospatial Business Model.

—Brian Timoney

End-of-year summaries and next-year predictions are the web’s way of helping you pass time during the most unproductive work week on the calendar. Or save you from continued contrived conversation among those with whom you share little except a similar genetic imprint. Rather than go the solipsistic blogger route and explain why The Decemberists put out the best album or that Incendies was my movie of the year, I’ve chosen a tack in which I’m more heavily invested. For “hotness” here refers not to PR buzz but tools that can solve both my clients’ current problems and their soon-to-be problems.

The next great GIS isn’t a “GIS”–but rather the statistical package R.  It’s the nexus where modelling, statistics, and graphics meet.  An open-source project with a large community and big developer momentum, there’s a critical mass of know-how such that you’d be hard-pressed to come up with a quantitative challenge that hasn’t already been tackled by the R community.  For mapping, the obvious starting point is the maptools package, but there are also hooks to familiar tools such as GDAL (RGDAL) and PostGIS, as well as the recently released GUI DeducerSpatial.

Let’s be clear, it isn’t about trying to replace your trusty GIS with statistical software. It’s about acquiring a more robust quantitative toolset to wrestle with a multi-variate world. Statistical clustering (spatial and non-spatial), principal components, multi-dimensional scaling, etc. will all be go-to techniques in a world that can no longer be explained by a single variable displayed on a map.  We all laugh at “red-dot fever” where lazy analysts overwhelm a map by displaying every coffeeshop, every bank, every whatever, creating visual confusion instead of anything approaching meaning. But the crashing of the tides of “Big Data”, the “sensor web”, and the “Internet of Things” upon our shores is imminent, and wrestling with those datasets with advanced statistical techniques will be the prerequisite for making meaningful maps.  A small taste of what’s possible in R is one of this year’s most compelling maps:  the Facebook map.

A lot of mapping shops will be scratching their collective heads this year figuring out how to serve a public that uses everything from Internet Explorer 6 to the iPad, as well as your preferred smartphone. With the mobile web leaving the worlds of Flash and Silverlight behind, where to turn for interactive vectors in the browser. Why not plunge into the future and go pure HTML5/SVG? Because the mapping community, with the large presence of government agencies at all levels has a disproportionately high use of older versions of Internet Explorer. Recent stats from ESRI suggest that visitors to ESRI.com use IE 6-8 at a rate roughly double that of the overall web user population.

Raphael is a javascript library that bridges the gap by rendering vectors natively as VML in Internet Explorer 6-8, and as SVG in the newer web browsers. Hence you get “live vectors”: rollovers, tool tips, click events, etc. without requiring plugins such as Silverlight or Flash.  Of course, rendering tens of thousands of vertices won’t go so well in older browsers, but for everyday thematic maps, such as this example of US States, it is perfectly serviceable. With 2012 being a Presidential election year in the US, expect many news organizations to move away from Flash to Raphael for their choropleth-ing of results.

Some of the most innovative work of 2011 came from Washington DC-based Development Seed. First, their open source TileMill cartographic studio is a much-needed tool that enables the making of visually compelling maps without GIS. This is a huge boon to the geospatial sector where too many of us, present company included, have been all too content to crank out always utilitarian, sometimes ugly, maps for our clients. Better still, the styling specifications use the CSS-inspired Carto language, making for easy re-use and sharing of styles.

And that’s not all. Using the very clever UTF grid approach, maps created with TileMill and served up via TileStream (also open source), feature interactivity that is also cross-browser–from Internet Explorer 6 to the iPad.  And because the interactivity is pixel-based, it can handle many, many thousands of features without killing your browser.

The coup de grace is the MapBox iPad app. All your great cartography created in TileMill with the full experience available in disconnected settings. In important ways, I see the iPad (and, hopefully, future tablets that can match the user experience) as a great second chance for mapping on the web. Because the first time around we as an industry failed our users by insisting on a desktop-GIS-inside-the-browser metaphor that was utterly foreign to anyone except fellow professionals. I’d love to see the default standard be a well-designed, informative basemap plus once “clickable” layer: enough information for 85% of your users without introducing the confusion of dozens of layers (tucked within layer “groups”!). And my observation tells me that there’s an intimacy users have with their iPad that is very different that their relationship to their desktop machines. Simple to use, informative, and aesthetically thoughtful is the big win here.

So the excitement about these new tools and capabilities would naturally lead to a conclusion that “it’s never been a more exciting time to be in geospatial.” But there’s also an underlying lesson in highlighting projects that don’t come out of the traditional group of GIS vendors: geospatial is attracting significant outside attention and people are getting things done using tools and methods that are unfamiliar to many industry veterans. Combine that with the contraction of the public sector that is a huge component of GIS employment, and the more sober conclusion is that our little niche traditionally off-to-the-side is more mainstream and much more competitive. So think about and spend time with some of these new technologies not for the “cool” factor, but to ensure the continued relevance of your skill set.

—Brian Timoney