[tei-council] Notes from today;s conference call

[Christian Wittern]

Sebastian Rahtz <sebastian.rahtz at computing-services.oxford.ac.uk> writes:

> Syd Bauman wrote:
>
>>>b) if Council does wish the META group to allow for this ability
>>>   in ODD, it's time to make concrete suggestions for a notation.
>>So soon?
>>
> yes. if it matters, it matters.

Well, the reason I brought this up was mainly to have it on record.  I
think we need to seriously consider the merits / de-merits of this
road, including proven need for it and its implications for the whole
of P5 etc., before we venture in this area.  It does not look like we
have the ressources for this at the moment, that's why I am very happy
to have this whole issue filed until we come across a place where we
actually need it.


>> While I very much hope that the future of TEI includes the
>>capability to constrain documents in manners more advanced than DTDs
>>would allow, perhaps even using the full power of RelaxNG, I think
>>META has very much earned a respite from continual work on the ODD
>>language, and that we (Council) shouldn't start pestering you (META)
>>with suggested syntaxes until you've had a breather. (The term "6
>>months" was bantered about during the call; I took this to be
>>figurative, but surely you deserve more than 6 days?)
>>
> the "6 months" I thought was "3 months"; but in any case
> what I meant by that was the the editors should have a guarentee
> of no more changes under their feet for the prescribed period. that does
> not at all preclude more work behind the scenes.
>
> one very very simple solution is to add an attribute to <attDef> saying 
> "this one's
> a dummy because its its been hand-declared in the content of the 
> element". But
> I don't want to go down that line of discussion unless the Council
> says we ought to. Which is why I wanted a quick discussion here first.
>

At the moment, honestly, I do not know the answer to this.  The reason
we pondered this particular case of co-occurrence constraint was that
some CDATA attributes might need markup itself, so would better start
life as a child element in some circumstances.  At the moment, I am
not sure if this argument still holds, since most of these CDATA
attributes should be gone now, thanks to the <choice> mechanism.  Lou,
Syd, do you have any idea of which attributes might still be in need
of such a construct?
 

All the best,

Christian

-- 

 Christian Wittern 
 Institute for Research in Humanities, Kyoto University
 47 Higashiogura-cho, Kitashirakawa, Sakyo-ku, Kyoto 606-8265, JAPAN