Klout is a service that purports to measure your influence across social media channels. The purpose of this post is not to render judgement on that concept but rather to show how to opt out. It was not a terribly intuitive process so I thought others might be interested in how to go about it.
I read with interest this article that was brought to my attention by @gis_Todd on Twitter. I have been playing with Foursquare for a few weeks now and I think this article is in sync with some observations I have had about it but also misses the mark on one key point. Continue reading “Foursquare – Teetering On the Edge of YASN”
Sean Gillies pointed out on Twitter today that the GeoJSON output of coordinates from Twitter’s just-released geolocation capability is wrong. Sean is one of the authors of the spec so, when I saw his tweet, I took notice. According to the Twitter documentation:
The XML response uses GeoRSS to encode the latitude and longitude. encodes as latitude, space, and longitude (see the response below for an example). For JSON the response uses conventions laid forth in GeoJSON which looks like
You will notice that this is in Latitude/Longitude order whereas the GeoJSON specification clearly states Longitude/Latitude.
This is clearly an error that will throw off any existing software that has been developed to handle GeoJSON. This is somewhat reminiscent of a similar issue with early versions of SQL Server 2008, with a lot of the same issues applicable here. The main difference this time around being that the GeoJSON spec is much clearer on X/Y ordering than the WKT spec was.
There seem to be a lot of clients already on board with supporting geolocation in Twitter so this could be a potentially large proliferation of this particular GeoJSON error. Twitter should look at correcting this before a lot of data gets out there.
Good catch, Sean. And, yes, it should have been noticed before now. Maybe we should look up from our iDroidBerries more often.
This past weekend, NFL Sunday to be specific, I was the victim of a TV/internet outage due to a mistake made by a technician while they did some work at the junction box shared by my next-door neighbor and I. As a result, I was unable to watch my beloved Washington NFL franchise hobble to another almost-loss so I had to find another way to keep tabs on the game. Since my internet service was also down, that left me with my BlackBerry Tour as my sole means of connectivity.
I initially started off on the ESPN site, trying to keep track of the action. Finding this unsatisfactory, I switched to MSNBC with the auto-refresh set to 30 seconds. I quickly realized that the information I was getting was woefully behind and quite inadequate. Then I had a different idea. Continue reading “Twitter Shows Blitz”
Ryan Sarver announced today the availability of a geolocation API for Twitter. It even supports GeoRSS and GeoJSON. This could be a potentially significant new step for Twitter. It looks like the API will be opened to
platform developers first everyone (see here).
Tying in location to the near-real-time nature of the Twitter timeline opens up lots of potential location-based applications. Let the slippy-mapping commence!
After being involved with zigGIS for a couple of years now, I’m excited that Paolo Corti is going to be visiting the States next week and that Abe and I will finally get to meet him (and vice versa). The three of us will be meeting at the Lolita Bar in New York City at 6pm on April 24th.
If you’re in the area (or you want to make a special trip), we could use all the help we can get to give Paolo a warm welcome to the US. I hope to see you there. Tweet us if you think you’ll make it.