Open Source GIS: Part II

I received a bit of email from my article last week on open source GIS. There are two letters from readers I’d like to share with you because I think they clearly represent two different perspectives of the open source GIS discussion.

I’d love to hear from more folks about their open source GIS apps. Please send me a quick email.

The first letter points out an important fact about open source GIS that I didn’t touch on last week. His point is that with open source GIS, anyone can modify the software source code to add or change features of the software.

Dear Eric,

I’ve been following your writings for a few years now, and I’m delighted to see you bring up open source geospatial software in your column!

Last year, I finished an online MGIS degree from Penn State, using open source GIS and web mapping tools with local township staff for my “capstone project.” Here in Michigan, townships are the local units of government that are below counties in the hierarchy of governments. There are over 1200 of them in the state.

Open source tools, I think, represent an “enabling” technology and movement, especially for impoverished township governments like those I worked with in northern Michigan. There, some of the staff may actually be trained in using commercial GIS tools (from previous employment or from schools), but their small, rural townships may not be able to afford licenses for them. In other townships, staff have enthusiasm for, but not training in, GIS tools, and are willing to spend the time to learn and use them if their townships can provide them with such tools. (Here again, up here in rural northwest lower Michigan, townships typically can’t.) The townships up here typically have populations in the low thousands.

For this segment of potential users, open source desktop GIS tools such as QGIS and uDig turned out to be just what was needed. For a couple of townships, I just had to show them how to download and install these tools, and then they were off on their own! After they created paper maps and such for their use (planning commission work, recreational planning, and so on), I made online interactive versions for their websites using open source tools. Such open source web mapping tools included Open Layers, Geoserver, and even Google Maps API, even though Google Maps API is not, strictly speaking, open source. In other townships, I initially had to sit down and coach them in the use of QGIS and uDig, but they soon caught on and started producing their own mapping products. Granted, the maps were fairly simple maps, but perfectly adequate to communicate to their constituents whatever needed to be communicated. Without open source GIS products, none of this would have been possible, especially in today’s economic climate.

If you take a close look at QGIS, for example, you’ll see that the current release version (1.7.4) and the development version (1.9.x) contain highly sophisticated tools that rival those from commercial packages. In fact, for the vast majority of potential users (such as those I worked with), what are in the current versions are more than enough for their needs. Through the efforts of the worldwide communities of enthusiastic, part-time developers who program in their own time, these open source geospatial products of great sophistication are made available to those who need them most!

Again, thank you for hi-lighting open source geospatial products in your column!

Howard Yamaguchi

P.S.  Eric, in your column you dwelled on the “free as in free beer” aspect of open source, where anybody can download the software and use it, gratis. The other aspect of open source that you could have mentioned is the part where, in addition to  downloading the product, we can also download the source code and tinker with it. We are then free to use the tinkered code for our own use, even to profit from it by deploying and using such modified versions for our clients. Open source licenses typically require us to eventually upload the modified source code to the developer community so that they can, if they so choose, incorporate it into the release versions of the product. That’s the “free as in free speech” part, the part that really distinguishes open source software from the world of commercial software. (But you probably knew all this!)  Cheers!

Following is a letter from a gentleman who says the cost-savings potential with open source GIS isn’t worth it for the organization he works in.

Eric—

Thanks for your always insightful columns. You asked for feedback from users who are using open source GIS apps, but might I offer the following counterpoint? The electric utility company I work for doesn’t use open source GIS and has no plans to use it, regardless of how reliable or low-priced it becomes. “Why?” you ask. “Is it because we’re herd-bound and lacking vision?  Are we so tied to our big-box traditions that we can’t see the coming GIS software wave?”

No, the reason is simply that the cost of neither software nor hardware is significant in comparison to the value of GIS data integration. When our system operators remotely close a switch, the distribution SCADA system’s electrical-connectivity data (which it receives from our GIS) darn well better correctly identify the facilities it just energized. So a GIS that’s highly integrated into our distribution-SCADA, accounting, work-management, outage-management and meter-data-management systems is a bargain at any price; but a stand-alone GIS app (even one with really great data-collection capabilities) is effectively worthless to us, regardless of how cheap it is.

This system integration is not something our management cares to do in-house (though I’m always making the argument that we can and should do more in-house), nor is it something we can entrust to volunteer hackers (regardless of how altruistic they are). There will always be a market for high-dollar GIS platforms that frictionlessly exchange real-time data with mission-critical systems.

It seems (to me anyway) that the only role open source apps might play is in areas that don’t involve interfaces with other mission-critical systems — e.g., in presenting our GIS data alongside other datasets, or perhaps in data collection/validation.

Thanks again, for your always interesting insights,

Bob

Thanks to Howard and Bob for their perspectives.

How about open source geo data such as openstreetmap.org?

If you recall, one of the trends identified by the experts that the United Nations polled was that “There is unlikely to be a market for datasets like those currently sold to power navigation and location-based services solutions in five years, as they will have been superseded by crowdsourced datasets from OpenStreetMaps or other comparable initiatives.”

Do you think that Bob’s rationale also applies to open source data? I’m sure that’s what Navteq/Nokia and Teletlas/TomTom would argue.

Personally, I think that the U.N. prediction of five years is pretty aggressive. Navteq and TeleAtlas have put a tremendous amount of time and effort into collecting and validating their datasets. They get paid to be accountable for the quality of their product. The challenge for OpenStreetMap to take over the GPS Personal Navigation and other LBS markets is that the quality bar has already been set. If OpenStreetMap data quality doesn’t match or come close, the user experience will be disappointing and OpenStreetMap’s reputation will take a hit.

Don’t take this wrong, though. Open source GIS data sources like OpenStreetMap, OpenSeaMap, etc. have a bright future. As I’ve written before, and as the U.N. correctly identifies, the proliferation of high-precision GNSS receivers, along with other sensors, will make high-precision data collection inexpensive and commonplace. Super-detailed, feature-rich, open source data is in our future.