On every day of all those everyday DH days, the TEI seems to be there… currently, I face three of its many faces:
- In the “Papyrus Project”, the digital reworking of the Cologne papyrus collection, it shows how sophisticated a TEI header can be when including manuscript descriptions for mostly fragmented papyri often bearing several different texts
- In the “Pessoa Project”, it shows how sophisticated a TEI body can be when representing handwritten sketches and lists with marginal notes, additions, substitutions, deletions, omissions, …
- In the “Computational Literary Genre Stylistics Group”, it shows how it can serve as a base format to hold essential metadata and a simple text structure, to be used in other (text based) formats with other tools
I have the impression that the more simple and straightforward you want the encoding to be, the more tricky it gets because of the many decisions to be made!