[DiapDealer]

This reminds me of the older "Adequacy of epub" thread. Which also boiled down to the fact that until a word-proccessing front-end exists (that is feature rich and intuitive enough that authors wouldn't dare to use anything else to create their masterpieces) which is capable of producing a semantic-rich and universally accepted--and standardized--markup representation of said work automatically, then this imagined future of auto-generated indices/compendiums and mined research data ain't happening.

People already don't use the semantic tools currently at their disposal with any sort of consistency. As such, I certainly don't see any advantage to a more complex and/or rigid specification that no two existing, industry-standard programs (which authors use to write/produce) would interpret consistently when producing their underlying markup. Nor do I see a new author-centric front-end capable of producing such capable, flexible, and perfectly semantic markup emerging to lure writers away from their current toolset.

Useful, flexible, industry-standard markup representations of books (novels or apps) are predicated on authors being able to produce such markup themselves. Either manually or with the assistance of their current preferred writing tools. In short ... any such future is a wholesale author paradigm-shift away. Until then, semantic-rich, highly-flexible, industry-standard markup is nothing more than a utopian pipe-dream.