Yesterday Yahoo released a new beta product called Yahoo Pipes. This tool has already been called “…a milestone in the history of the Internet” by visionary Tim O’Reilly. The full concept may be a little tricky to grasp if you are not a technologist, but in a nutshell it enables you to easily take the output from one Internet site and feed it to another site as an input which then lets you do something new with the data. (Those of us who have lots of Unix experience are quick to grasp the concept). The Google Maps mashups you hear about all the time (combining unique data and applications with maps) are an example of manually developed “pipes”. Yahoo Pipes makes it even easier using a graphical diagram to architect by drag and drop mashups (and more). So, for example you can take the home page of the New York Times, find all the words which are locations, and then find pictures at Flickr which represent those locations. What I am excited about is that it should be possible to include applications which can handle KML and both process and generate GE content combined with other data sources to generate really unique GE visualizations. Read on for an example of what I’m talking about in the form of a thing called Earthify.
Google Earth already has some aspects of this in the form of the network link. A network link lets you get the KML from another server somewhere which may tap into other data to provide you with dynamic content in GE. For example, the near-real-time NASA photos from their MODIS satellites overlayed into GE. Combining Pipes with network links will make for truly powerful things.
So yesterday someone released their own project which demonstrates the possibility of pipes (although he didn’t use Pipes, and was in fact thinking his efforts were wasted now that Pipes is out). Jeff Crouse released Earthify, which takes the results from a page in Craigslist and maps those results into Google Earth. His application scans the Craigslist results and determines which entries are “Earthifyable” (where he can determine a location based on addresses), and which are not. Then sorts the list and generates a KML file to show the results in GE. Visit this page for both an explanation of the application and instructions on how to set it up. In his case, he provides a bookmarklet which lets you easily process a Craigslist page (thanks to the code he wrote). Jeff implemented this so he could evaluate housing choices based on viewing the locations of Craigslist listings. Great work Jeff!
Someday soon, with the advent of Pipes, we will see very cool GE visualizations like highly personalized weather forecasts where the data could be taken from various sources (NASA, FAA, NOAA, weather.com, and your local weather channel) to provide you with more meaningful information all by just turning on a network link inside GE. You could already do this with a bit of programming, but the Yahoo Pipes concept will make it much easier. This is going to be great!
About Frank Taylor
Frank Taylor started the Google Earth Blog in July, 2005 shortly after Google Earth was first released. He has worked with 3D computer graphics and VR for many years and was very impressed with this exciting product. Frank completed a 5.5 year circumnavigation of the earth by sailboat in June 2015 which you can read about at Tahina Expedition, and is a licensed pilot, backpacker, diver, and photographer.
Haven’t tried pipes yet since the site was down, but it seems that this is mostly a tool to allow non-programmers to do programmer like things using a visual programming tool. So in that aspect it will be good to have more data feeds, but I’d still like to see some convention/standard xml schemas or controlled vocabularies in regards to data content( http://nautilus.baruch.sc.edu/twiki_dmcc/bin/view/Main/ObsKML ). If I do a google search on “Frank Taylor” am I talking about the google earth blogger or the sales agent to the dental profession? Without some manual oversight and work, my scripts can’t distinguish. Individuals will continue to publish some remarkable feeds but its all geared towards individual consumption and missed gains in continued recombinations at the machine-to-machine level.
Earthify is great!! Where can the address be translated to the latitude and longitude values in order to create the KML file?
Jeff Crouse here – I made the Earthify script. It uses the Google Geocode API – you can find a tutorial on it here:
http://www.developer.com/db/article.php/3621981