[tei-council] handShift anomaly

[David J Birnbaum]

Dear Lou (cc Council),
> But doesn't this miss the point which James noted?  sometimes it is 
> *useful* to have both @new and a child <desc> -- not that the <desc> 
> contradicts or duplicates the info provided by the @new, but that it 
> qualifies or nuances it. I think that requirement is one which can 
> reasonably be specified in the prose.
I think I understood James's point, but I'd be inclined to view a 
nuanced or qualified value as a different value, rather than a subtype 
of an existing one. For example, I wouldn't treat dark green as a 
subtype of green, with the greenness to be specified in the @new value 
and the saturation to be specified in <desc/>. I'd make them different 
@new values.

I can live with allowing both @new and <desc/>, but since allowing only 
@new doesn't prohibit nuancing or qualifying, but, rather, just requires 
that it be done in a way that doesn't create an opportunity for 
contradiction or duplication, I think it's likely to be safer.

But let's get rid of @old in any case.

Best,

David