> > > would it satisfy this need if revisionDesc were allowed as a
> > > child of <div>? its one way to go.
> > That's the one I'd like to see.
> does any one else support this? it seems curiously attractive
> to me.
My gut reaction was to say "catagorically not". I think it is a very
bad idea to mold our encoding methodology to meet the restrictions
imposed on us by current tools, rather than wait until better tools
come along (perhaps with our encouragement). And certainly I don't
think we would allow <revisionDesc> inside anything other than
<teiHeader> here at the WWP, as it breaks what we consider a
fundamental distinction -- metadata in the <teiHeader>, marks that
were on the page of the source in #PCDATA inside <text>, and
information about those marks and their meaning in the tags inside
<text>.
On the other hand, just because *we* wouldn't do it, shouldn't close
my eyes to the possibilities. And there is something to be said for
being able to directly indicate which part of the document tree has
been altered[1]. (But if <revisionDesc> could be a child of <div>,
why couldn't it be the child of <text>, or <lg>, or even <list>, <l>,
<item>, or <p>? Heck, why not of <name>?)
Notes
-----
[1] As opposed to what I would call the "normal" indirect method:
<change corresp="N-04">
<date value="2002-05-20"></date>
<respStmt><name>Syd</name><resp>too much</resp></respStmt>
<item>Added new editorial note about ducks.</item>
</change>
<!-- ... -->
<p>XML is ubiquitous.
<note id="N-04">Ducks, it turns out, do not use nor apparently
even appreciate XML. This based on some observational studies
performed at a nearby pond. I am anxiously awaiting reports
from a sizeable number of well-renowned zooligists,
biologists, and naturalists about how other members of the animal
kingdom approach XML.</note>
...</p>
|