>I had not realised, until I read that, that <note> did not allow notes
> to span multiple paragraphs, or contain lists, quotes, etc.
[And a snip from an off-list posting from Michael]
> I suspect you've [...] fallen foul of the deceptively user-friendly list
> of elements at the end of P4. This is a danger zone [...] The <note> you
> are both describing looks to me like the <note> used (or rather not used,
> see below) in WSD's.
No error as such, although I should have been clearer in what I was
The "main" <note> has a content model of %specialPara; although this allows
a considerable flexibility in terms of chunk and inter-level elements, so
Steve you certainly can include all kinds of <p>, <list>, etc. elements. It
just doesn't quite extend to the full <div> structure (mainly it does not
include %divtop; and %divbot;, or permit nested <div>s), which is what I
understood the original questioner to be after.
> Which, for most of the footnotes in the works I've been looking at, makes
> <note> completely useless, and I'm starting to wonder what the TEI gods
> smoking when they designed this thing.
Go ahead and use note as above. As Michael says, make sure you're looking at
the right page in the element reference. It does pretty much everything
you'd expect except actual <div>s. Which is fair enough, in my view.
> The guidelines actually suggest that you embed footnotes in the point at
> they are referenced, but that's not possible, using <note>, for anything
> than very simple footnotes.
> I guess the alternative is to place the footnotes in a separate <div>, and
> <ptr> or <ref> at the point of reference. Which was my initial impulse
This is what I was essentially suggesting for the original questioner, but
they did desire the full structure of a <div>. If you've got relatively
ordinary footnotes, go ahead and encode them with <note>.
>>>Should you ever try to place a <p> after a sibling-level <div> then
>>>the inherent order of the universe will be disrupted, and Heaven help
> Also wrong in my opinion, and NEEDS TO BE FIXED! (And I've read the
arguments already, thanks.)
They've been pretty clearly stated on all sides. I've long since come to the
conclusion that it's an almost metaphysical thing ;-) There are those who
believe that there is such a Thing as a <div> (in abstraction), and that one
of the essential attributes of a real-world instantiation thereof is no
orphanage; and there are those who belive that <div> is just a name we give
to a bunch of similar phenomena which may, or may not, have an essential
attribute of orphanage. Although I certainly belong to the latter camp, as I
mellow into my old age I am coming to the opinion that although I believe
that I should be *allowed* <p/><div/><p/> it certainly shouldn't be
*encouraged*, and as such maybe it's acceptable to leave it as it is, and if
someone really wants it they can use the TEI extensions mechanism to allow
it. Or in P5 we could define both patterns in the modules, and include a
switch to select.
 http://www.tei-c.org/P4X/ref-NOTE.html for the definition of (the
"real") <note>; in particular:
May contain: #PCDATA ab abbr add addSpan address alt altGrp anchor app bibl
biblFull biblStruct c caesura camera caption castList cb certainty cit cl
corr damage date dateRange dateStruct del delSpan distinct eTree emph expan
fLib figure foreign formula fs fsLib fvLib fw gap geogName gloss graph
handShift hi index interp interpGrp join joinGrp l label lang lb lg link
linkGrp list listBibl m measure mentioned milestone move name note num oRef
oVar orgName orig p pRef pVar pb persName phr placeName ptr q quote ref reg
respons restore rs s seg sic soCalled sound sp space span spanGrp stage
supplied table tech term text time timeRange timeStruct timeline title tree
unclear view w witDetail witList xptr xref