I've got a question about since_id and if it actually works.
The API states -
since_id: returns tweets with status ids greater than the given id.
So when I get a list of 100 twitter messages, I store the maximum id
in that group, which I get from the id property after stripping out
"tag:search.twitter.com,2005:" and use that id for my since_id the
next time I run the query. The problem is I've noticed I get back id's
that I've already seen in previous requests when using since_id. It is
fine for me I just ignore them I just thought I'd bring it to the
attention of you guys, unless I've misinterpreted this functionality.