feoNotes

Notes and Reflections by feoMike

thoughts on disruption

earlier this week, GitHub released a very small piece of code which does one small specific thing; it recognizes a geoJson file in any GitHub repo, and automatically places a mapbox map under those tiles with simple pan, zoom and identify functions in a web map. it is a remarkably simple feature, which i think significantly disrupts the traditional geospatial market. the end result of this tiny enhancement to GitHub, will, like many disruptions, significantly change the landscape of traditional geospatial practitioners. here's why.

the traditional geospatial market, predominantly government in my experience, but industry verticals too, use geospatial software for two primary purposes. purpose one is to maximize the full suite of data management and analysis utilities in order to research and investigate something on the landscape. these tools can be remarkably complicated involve large data and even require an advanced individual with very specific training to perform the analysis, back end software, data infrastructure, and computer network setup. most often these tools allow a skilled individual to look at, evaluate or otherwise manipulate the spatial relationship in these data set; is this feature close to that one, inside this jurisdiction, along this line/path, a neighbor to this feature or this far away from this one. while relatively simple sounding, this manipulation is hard, computer and personnel intensive and often costly. In many cases, this literally involves rocket science, physics, computer scientists, or ecological principles to name a few.

purpose two is to communicate the results of purpose one; essentially place colors, shapes and labels of certain analysis on top of a basemap. the design of these two items becomes critical. the basemap gives us context of things we already know. i can usually recognize political boundaries (like states and counties), man-made infrastructure (like roads and buildings) and physical features like streams and mountains. usually we have seen these so many times, we subconsciously know exactly where we are without having to really spend time on the map. for instance we know that washington dc is halfway between maine and florida, that the sierra nevada range is somewhat coincident w/ the california nevada border, and its relatively flat between denver and say eastern ohio (ok not really, but). the results are then placed on top of the base context and highlighted; the red color here shows a creeping fire, or lots of environmental issues in this area, or too much traffic on that route home. imagine if we had to read text descriptions of all these complicated issues. compare that with the last severe weather map you were looking at. the later gives the consumer instant knowledge of the issue.

communication, should be easy. i should be able to take data from research, and place it context of what we already know with little effort. i shouldn't need the advanced software, a platform or some other credit based service. moreover, anyone should be able to consume that result and not have to have a phd in some field in order to consume the map. i should be able to consume that map, anywhere, anytime and without software, hoops, logins, credentials or training of any kind. i should be able to do it in a browser, with the most common things i am used to; recognize the basemap context, know how to move around on a map, zoom in/out and point at something with a cursor and have the computer tell me it is this thing (like a bike station).

well, GitHub made it SUPER easy. they made something that is actually part of the web ecology. they changed the game entirely, radically, and as i understand it with a crazy few lines of code. i have calls every day from people i work with who say, can i get a map of just x (the results of a question). now i can return that as a url to them with near 0 production effort. step 1) place results of question in geoJson file here, step 2) send someone link to repo.

done. no software engineers involved. no database administrators. no network engineers. no development designers. no procurement of any services. no contracting officers. no logins, bulky software, or even really custom tools.

will this work for every situation? no absolutely not. but it will work for perhaps 80% of mine, and maybe 100% of many. i am quite confident GitHub will build on this to make it even easier, and they will further disrupt the market. i know we will be watching and take advantage. perhaps we already have or better yet, this.

GitHub is disruptive because the next disruption will just build on this, from some kid who codes and makes the world a better place. and there will be many many more.