[tei-council] Draft Roma Replacement Specification

[James Cummings]

On 10/11/13 13:50, Sebastian Rahtz wrote:
> it's <specGrp> and <specGrpRef>  that make this tricky, I think. not impossible, but I never dared contemplate them in Roma

True, hence why I don't think it should be a requirement. (But if 
someone implementing it wanted to say have a way to push out to 
something like eXide to enable someone to edit it... I wouldn't 
mind it.)

> also the markup within <remarks> (and to a lesser extent <desc>) is of course open-ended.

True, so the requirement should be that someone can enter that 
data _somehow_ in NuRoma but that this shouldn't necessarily be 
unconstrained. Or that if markup is there it shouldn't be 
discarded.  I think as a general rule that if I load up an 
existing ODD and change one element, then the whole rest of the 
ODD should remain unmolested if possible.

-James



-- 
Dr James Cummings, James.Cummings at it.ox.ac.uk
Academic IT Services, University of Oxford