Well, when Twitter changed their API rules to survive the summer of FAIL whale, it crippled the service. And, as such, we’ve been scrapping and scraping to find resources that allow us to find local folks.
Never let anyone say that Matt King won’t find a better way.
Since Twitter cut off their Jabber feed from TwitterLocal, we had to rely purely on the XML API, which meant that only about 20% of Tweets from the public timeline got into TwitterLocal. Now that Twitter has a location-based search API, we don’t have to cache the posts anymore. So now, TwitterLocal is going to be purely an Adobe AIR based application that allows you to filter Tweets by location.
With the new Air app, TwitterLocal supports regional searches with various radii. And it supports multiple tabs—so that you can watch a number of different regions at the same time.
Plus, the beauty of AIR is that it’s crossplatform, right out of the box.
The leader board currently ranks cities by the number of tweets by residents in a rolling 24-hour period.
Glancing at it a few minutes ago, Tokyo was in the lead with San Francisco running a close second. Paris leads the Europeans. And our hometown of Portland is sitting around #14 or so.
From 8:00PM, April 8, 2008 through 8:00PM, April 9, 2008, the list looked something like this:
New York City
(Entre mi cuarto y mis zapatos)
(Mexico Distrito Federal)
As you can see, there is some weirdness can show up in the results. King notes these flaws in the system:
The seemingly high count of random places like “my pc”, “cybertron”, etc. are the geocoding service’s way of having fun. It seems some fake locations get assigned coordinates to somewhere in Kansas.
There is also a very high count of locations with asian characters, which again the geocoding services give only one location. Other than that the numbers are fairly accurate.
Despite these minor foibles, TwitterLocal’s leader board is the first location-specific Twitter analysis that I’ve encountered which actually begins to show which locations have caught the Twitter bug.
And as impressed as I was with TwitterLocal’s service, I’m sure to find this type of competitive ranking completely addictive, at the very least. I’m sure I’ll be checking TwitterLocal leader board, obsessively, over the coming months to see if we can get Portland to crack the top 10. At the very least.