I've got frustrated of selecting the right licenses in many of my projects, because of the subtle differences in all of them and the large incompatibilities while mixing them. Also, I often find that some part of a document should be more restricted that the rest (think of a documentation you want in public domain citing GPLv3 code, or a GPLv3 code citing restricted standard text). Also, this scheme should be able to merge both "textual" licenses with "artistic" licenses like CC.
Hence I'm proposing a new scheme that clearly specifies what "rights" you're reserving for what part of a document, media, opus, speech. It's composable, it's simple to use for both humans and machine, it allows to deny machine processing (under the jurisdictions that do protect citizens against those).
I'm asking for feedback for this idea, is there something that you'd like to protect that I've forgotten, or do you find it convenient to use and understand ?