Re: [docbook] RDFa Lite inside <info>

[Peter Flynn <peter@silmaril.ie>]

On 03/07/2021 10:21, Dave Pawson wrote:
[...]
> What's the logic for making metadata available in HTML please?

To allow it to be harvested by indexing engines to make pages easier and
more accurate to find.

(Hollow laugh)...if only indexing engines payed any attention to it.

On 03/07/2021 11:30, Dave Pawson wrote:
> On Sat, 3 Jul 2021 at 10:35, Norm Tovey-Walsh <ndw@nwalsh.com> wrote:
>> I tend to think of metadata as being about the document [and] 
>> annotations as being commentary for human readers.
> I'd suggest both are (eventually) for human readers, and both are
> 'about' the document?

Eventually, yes. If I add to my book (for example) ddc="004.67" most
humans will need a librarian to resolve it (to computer science: wide
area networks) because that's a job usually done by a cataloguing
system. I would suggest that that degree of resolution is out of scope
for the markup: the Dewey metadata is sufficient.

> Why would I annotate book A with information about book B?

Probably because they're in the same series, or you need to read one
before the other. Like publishers used to list an author's earlier
publications on the half-title verso.

Apart from title and author, IMNSHO metadata is often just tokens (like
the name="value" above) whereas annotations are more often more like
sentences.

I have often abused the annotations attribute for things like "this
worked in dev and I'm damned if I know why it fails in prod".

Peter