In the HTML5 mailing list, I'm proposing allowing use of more compact
syntax for semantics:
So, for representing:
...instead of being forced to use the following more awkward construct
(itemid ("namespace") for TEI defined earlier in the document for
<meta itemprop="agent" content="water"/>
...one could just use:
<span itemprop="damage" item-agent="water">
...offering the additional advantage of not requiring data-* attributes
to allow targeted styling of water-damaged elements.
However, one of the community members is making the argument that they
would like to make creation of hidden meta-data more difficult even
though they recognize it's need sometimes.
I'm planning to respond in demonstrating how hidden meta-data can be
essential to benefit from the richness of the likes of TEI, but was
hopeful of having a list of examples of how TEI semantics could be
targeted in (search engine) queries and transformations such that an
average person could understand their utility. In this case, "hidden
meta-data" translates really to TEI attributes, since in HTML using them
would need to be done as hidden meta-data, but I'm reall ytrying to
showcase what TEI semantics offers in general.
My examples so far include:
* Finding all of the letters whose openers begin after a certain date
* Finding all of the unique closings of a letter writer.
* Finding all of the quotations attributed to a particular person
I was hoping others might be able to share common or interesting use
cases (esp. not possible in HTML, but if you don't know HTML, please
share them anyways), since so many decisions in the HTML5 group are
based on the use cases.