[tei-council] missing refLists (was Re: <g> in <line> not allowed)

[Lou Burnard]

On 19/04/13 11:08, Sebastian Rahtz wrote:
> On 19 Apr 2013, at 10:21, Lou Burnard <lou.burnard at retired.ox.ac.uk> wrote:
>
>> On 19/04/13 10:12, Sebastian Rahtz wrote:
>>> i am wondering whether to automate this by looking for the
>>> nearest ancestor <div> which has a specDesc pointing at the relevant
>>> element. would people agree this is a worthwhile strategy?
>> we probably need references to *all* such <div>s, since <specDesc>s for the same element can appear in multiple places.
> fair does. if there's specDesc for an element, link to it
>
> mind you, i do think putting it in the source is redundant. if it can be derived automatically, which fossilise the current state? if you add a new
> section to a chapter and add a specDesc pointing at <line>, you''ll probably forget to update the spec for <line> itself


That looks like the logic by which we stopped putting them in. With the 
consequences we now see.