The problem seems to be how to measure such an effort. It's easy to list a few classes and subclasses (of the patent database) and say that you've searched thro

The problem seems to be how to measure such an effort. It's easy to list a few classes and subclasses (of the patent database) and say that you've searched through those entirely, but the whole internet? What documentation would be sufficient to convey that the Examiner had searched thoroughly enough? Examiners are already urged to search NPL (non-patent literature) sources, but it seems too amorphous to establish standards for what sources need to be searched and how thoroughly. And the tooling isn't there either. Natural language search engines aren't precise enough for a real search for information (don't want the googles guessing what you 'really' meant to search for). When I was an examiner, for what it's worth, I always searched at least IEEE, CiteSeerX, and Dialog databases, but I'm not sure that a KickStarter campaign would have come up in any of those databases either.