[tei-council] invalid examples allowed in Guidelines?

[Kevin Hawkins]

On 1/9/2011 12:16 PM, Sebastian Rahtz wrote:
>
> On 9 Jan 2011, at 17:10, Kevin Hawkins wrote:
>
>> Related to this, Syd Bauman is currently working on creating ODDs for
>> the encoding levels described at http://purl.oclc.org/NET/teiinlibraries
>> and has discovered that you can't include @xmlns on any elements within
>> the examples.
>
> Of course you can. Whether its valid or not depends on the schema you build
> to validate the ODD.

I've asked him to get in touch with you to explain.

>> To clarify Lou's idea while revealing my ignorance, would it be
>> sufficient to put egXML into another namespace, or would every element
>> within an example also need to be?
>
>
> if you said<egXML xmlns="http://www.example.com/foo.bar">
> then the<egXML>  itself would be invalid, and the stylesheets
> would have to extended to know what to do with this new beast.
>
> the elements inside<egXML>  are normally in the same namespace
> as the<egXML>  anyway if you use @xmlns (its value is inherited)

I stated my question sloppily, leading to confusion.  Let me try again.

Lou wrote, "I wonder whether declaring them to be in some other non-TEI 
namespace would be appropriate."  I'd like to clarify Lou's suggestion 
by asking the following:

A) To avoid validation errors at build time, would it be sufficient to 
redefine the element egXML so that it is in another namespace besides tei:?

B) Unlike Lou's solution d (convert all these class of examples to CDATA 
in <eg> so they are not checked at all), redefining the namespace of 
egXML lets us continue checking for well-formedness of examples. 
However, am I correct to understand that by doing this, we would lose 
the ability to validate examples at all?

Kevin