Lou Burnard wrote:
> The idea was to make all listXXX elements  have similar content models, 
> but listPerson seems to have been left out. The standard model would 
> give listPerson a model like this:
> headlike*, (personLike|listPerson)*, relation*

that looks very good to me, much better than stuffing the relations into 
a wrapper element. Is this change likely to go into P5 1.0?

> i.e. you would nest one listPerson inside another rather than 
> arbitrarily call the outermost one a particDesc.

The one advantage of <particDesc> appears to be that it may contain 
'model.pLike'. We don't use this, but others might find it useful.
Therefore, the model would probably be something like (model.pLike+ | 
(headlike*, (personLike|listPerson)*, relation*)). If the content-model 
of <particDesc> will be equal to the model of <listPerson>, then one 
could probably drop <particDesc>.


cheers,

Stefan

-- 
| Stefan Majewski  | Department of English, University of Vienna     |
| VOICE Corpus     | Spitalgasse 2-4, Universitätscampus AAKH, Hof 8 |
|                  | A-1090 Vienna                                   |
| Research Ass.(IT)| Phone: +43 1 4277 424 46                        |