Eventually, documents written in, for example, DocBook will become
sources to researchers in the humanties (like historians of science and
technology).
Do we need a more record like text encoding, used for the automatic
indexing, searching and retrieval of heterogeneous collections of
documents?