For each binary blob of a hypertext Web page (images are the most common such blob, but video, flash, PDF, embedded objects, and even HTML, Javascript, and CSS could be), calculate a cryptographic hash of the data. Then, at the user's request, do a web search for the hash value. The search results allow for external commentary on the image or blob, external to the site itself, and free (or freer) from any censorship the site imposes.
If you want to comment on an image, you post on your own blog (or something like Usenet) the hash and your comment. A standard format of comments, ratings, thumbs up/down, would be useful for aggregating many search results.
We would like to avoid the privacy leak of the search provider knowing what images you viewed. Use tor.
One might hash the decoded image data, not the compressed original in order to catch different versions of the same thing.
Hashes of Javascript might be useful for collaborative filtering of whether a script may be trusted to run.
Another blob might be the FEN string describing a chess position.
This mechanism actually frees the site operator from having to run a forum comment board, with all its political difficulties of spammers, trolls, and obscenity. Essentially PageRank is used to filter or rank bad comments. I'm not sure how good it will be.
Like UUID.
No comments :
Post a Comment