They coalesce into a digital harmony
August 10, 2017 7:52 AM Subscribe
RandomCity is a project by Patricio Gonzalez Vivo, an artist and engineer who uses code and light to turn data into stunning landscapes. More of his projects here.
This link opens on Lower Manhattan. You search your own location on the upper left.
This link opens on Lower Manhattan. You search your own location on the upper left.
http://patriciogonzalezvivo.github.io/tangram-sandbox/tangram.html?styles/tilt-ikeda#17/40.70328/-74.00666
You don't say...
posted by acb at 8:48 AM on August 10, 2017 [1 favorite]
You don't say...
posted by acb at 8:48 AM on August 10, 2017 [1 favorite]
I am reminded of William Gibson's description of cyberspace as resembling a cityscape, with concentrations of data resembling buildings (though not necessarily rectangular, as with the ziggurats of the Eastern Seabord Fission Authority that Maelcum interprets as Babylon).
posted by Gelatin at 8:59 AM on August 10, 2017 [1 favorite]
posted by Gelatin at 8:59 AM on August 10, 2017 [1 favorite]
This is a great example using Tangram, a 3d/openGL open-source map rendering library (rendering: taking the geographic data - points, polygons, etc and turning it to 3d or 2d graphics) library by Mapzen; Patricio is one of the lead Tangram lead developers.
I've been experimenting with Tangram on/off for the past year or so; The learning curve is extremely sharp (I had never done C or anything like it before) but it has some capabilities (writing your own custom shaders and a few more that I'm forgot) that the main open-source competitor (MapBox GL)was missing but MapBox is gaining ground on the technical aspects and also released a Unity library (I'm unclear on the differences between Mapbox GL & Unity) is getting more attention from developers & businesses than Mapzen's Tangram.
As someone who's dabbled in online cartography and been on the sidelines of watching both of these companies and their capabilities grow, and have used both companies' offerings, I'm interested to hear others' opinions why/how MapBox has grown faster than Mapzen although that's a derail...).
The maps' accuracy vary greatly because nearly all of the source data is from Openstreetmap. In most cities, building heights are likely missing in (NYC, SF, are two exceptions that I know of in the USA at least; coverage in Europe is very generally a bit better; and varies for the rest of the world); and many cities are missing buildings, let alone the buildings' heights.
If you want to see more examples of his work and the code, check out the sandbox which features several more rad 3d experimental designs.
(also, a cheap plug, here's my repository of simpler snippets of code (called scenes) that you could use to make your own maps with Tangram; I still haven't been able to completely nail down 3d buildings just yet.
posted by fizzix at 9:04 AM on August 10, 2017 [2 favorites]
I've been experimenting with Tangram on/off for the past year or so; The learning curve is extremely sharp (I had never done C or anything like it before) but it has some capabilities (writing your own custom shaders and a few more that I'm forgot) that the main open-source competitor (MapBox GL)was missing but MapBox is gaining ground on the technical aspects and also released a Unity library (I'm unclear on the differences between Mapbox GL & Unity) is getting more attention from developers & businesses than Mapzen's Tangram.
As someone who's dabbled in online cartography and been on the sidelines of watching both of these companies and their capabilities grow, and have used both companies' offerings, I'm interested to hear others' opinions why/how MapBox has grown faster than Mapzen although that's a derail...).
The maps' accuracy vary greatly because nearly all of the source data is from Openstreetmap. In most cities, building heights are likely missing in (NYC, SF, are two exceptions that I know of in the USA at least; coverage in Europe is very generally a bit better; and varies for the rest of the world); and many cities are missing buildings, let alone the buildings' heights.
If you want to see more examples of his work and the code, check out the sandbox which features several more rad 3d experimental designs.
(also, a cheap plug, here's my repository of simpler snippets of code (called scenes) that you could use to make your own maps with Tangram; I still haven't been able to completely nail down 3d buildings just yet.
posted by fizzix at 9:04 AM on August 10, 2017 [2 favorites]
Gelatin, that was also my first thought. Neuromancer was set after a limited nuclear war in Europe between the US and the Soviets. Maybe in reality it will be a limited nuclear war on the Korean peninsula that results in the erosion of normal government power and the ascendancy of the megacorporations as the true governing powers.
posted by Sangermaine at 9:20 AM on August 10, 2017
posted by Sangermaine at 9:20 AM on August 10, 2017
Help me understand: what "data" is being turned, by code and light, into these landscapes?
Are we seeing a realtime representation of a 'live' data stream?
posted by armoir from antproof case at 10:38 AM on August 10, 2017
Are we seeing a realtime representation of a 'live' data stream?
posted by armoir from antproof case at 10:38 AM on August 10, 2017
I had the same question as armoir, I would love to know what specifically is being interpreted. This is a really interesting project, regardless, but I'm curious about that piece.
Also, heads up that this site might explode your browser history. I'm on Chrome on Android and I have pages upon pages of history from this site after just zooming out.
posted by motioncityshakespeare at 11:13 AM on August 10, 2017
Also, heads up that this site might explode your browser history. I'm on Chrome on Android and I have pages upon pages of history from this site after just zooming out.
posted by motioncityshakespeare at 11:13 AM on August 10, 2017
Armoir and motioncity, the data is from Openstreetmap (OSM) (a global source of geographic data with minimal licensing restrictions, very hastily: think 'google maps' but with the liberal licensing and crowdsourced nature of wikipedia ).
The data from OSM consists where certain permanent objects are situated (roads, businesses, lakes, buildings, etc) are situated but also its properties (the road's name, # of lanes, kind of road; or the # of floors the building has, etc).
Regarding Openstreetmap, the data display in the map is likely refreshed from there pretty often (my guess is every few hours, likely not longer than a day) to account if there's any changes (but people watch for vandalism, etc).
The little blobs on the highways or buildings likely aren't representations of any real-time data; they are just randomly generated groups of pixels intended to convey an aesthetic.
The rotations of the map's perspective when you zoom appear to depend on the time of day and if your device is moving (just a hasty read of the source code; but I'm a bit rusty at JS and am not familiar with those functions); but I'm not completely sure how much it (time and device moving) is influencing the map's perspective changes.
/ not the author but I sometimes make maps using the same data and software as author who does a hella better job than me.
posted by fizzix at 1:02 PM on August 10, 2017 [1 favorite]
« Older It Belongs in a Museum | We jumped the shark in, like, Episode 2 Newer »
This thread has been archived and is closed to new comments
posted by carter at 8:02 AM on August 10, 2017