[tei-council] rowing back on <mixedContent>

[Martin Holmes]

On 14-07-04 11:18 PM, Sebastian Rahtz wrote:
>
>>
>> Oooh, good question. I think it has to, doesn't it? You're
>> insisting on the presence of something (small t) before
>> someThingElse, and if there's nothing which could be there, then
>> there can't be any valid uses of the element. If a schema includes
>> an element which cannot be used, then something's surely wrong.
>>
>> I guess what should be flagged by an ODD processor would be the
>> inclusion of a model which has no members. That's unambiguously
>> wrong, isn't it?
>
> Indeed. Out the window go many customizatiins...

Should it just be the default behaviour of an ODD processor to remove 
any model that has no members from the schemaSpec wherever it occurs?

Then you could end up with content models which have no content, of 
course. But that might arguably be what you want, so it's not 
necessarily wrong.

Cheers,
Martin

> Sebastian
>