Hi all, I dev for an application where we do timely domain-based searches for all recent tweets involving that domain. Part of the queueing process is automated, and has been working brilliantly, until i discovered an odd quirk just this morning.
Let's say a tweet was made linking to www.example.com/one/long/uri (shortened via t.co), and i do a search: http://search.twitter.com/search.json?since_id=1111111111111111&q=www.example.com&rpp=100 That search will fail to find the tweet in question. But this will succeed: http://search.twitter.com/search.json?since_id=1111111111111111&q=example.com&rpp=100 Having noticed exactly this (on a real client's domain), i looked through our DB to notice that no searches on domains prefixed with www. domains had ever succeeded, and a quick sampling showed that tweets had been made on those domains, provided i excised the 'www.' from the search term. I couldn't find this behaviour documented anywhere, and i'm wondering - can it be trusted? If i start stripping www. from my search terms, is it a reasonable expectation that tweets linking to the www. subdomain (and shortened versions of those links) will continue to show in the search? Thanks, Phil -- Twitter developer documentation and resources: http://dev.twitter.com/doc API updates via Twitter: http://twitter.com/twitterapi Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list Change your membership to this group: http://groups.google.com/group/twitter-development-talk