Filtering and identifiers and logging.

[Cantor, Scott]

On 6/21/13 6:07 AM, "Ian Young" <ian at iay.org.uk> wrote:
>
>I think the schema is broken, whichever way you look at it.  The
>intention is clearly to have a "ref" attribute holding the reference, but
>there's not "use" specified so the "ref" ends up optional.

That's true.

>  And yes, the code looks as if it reads the text content rather than the
>ref attribute but the schema doesn't say anything about the element
>content.  I guess it's another of those cases where you end up with an
>open content model by accident.

No, it's not open. There's no sequence, and no base type, so it's closed,
no text allowed. I verified that. So we basically have a choice as to what
to do for V3, because there's no way to make the schema syntax work, and
no way to get it to accept the syntax that does work.

>Of course, as soon as I learned about it I thought "way cool" and wanted
>to configure my ARP using it heavily.  I guess I should, um, not bother
>with that.

Apparently not.

-- Scott