I used to be able to grab the "refresh" link out of an xml document returned from a query like http://search.twitter.com/search.atom?q=iphone . however, now, after two iterations of grabbing the refresh link, I get 403s back from search.twitter.com. the refresh link appears to be broken/poorly encoded.
steps to reproduce via curl: request 1: http://search.twitter.com/search.atom?q=iphone - works fine request 2: http://search.twitter.com/search.atom?q=iphone&since_id=5393893759 - this url came from the refresh link within response body's xml document returned in request 1. - works fine request 3: http://search.twitter.com/search.atom?amp%3Bsince_id=5393893759&q=iphone&since_id=5393907862 - this url came from the refresh link within response body's xml document returned in request 2. - results in a 403 from the server.
