Re: ISSUE-97 (metadata content model): Poor specification of metadata content model [DFXP 1.0 (Editorial)]

I think it is easier to handle this by saying that ttp:profile only has
semantic effect if it is a child of tt:head; we use similar language for the
ttp:* attribute vocabulary

On 5/15/09 5:44 PM, "Timed Text Working Group Issue Tracker"
<sysbot+tracker@w3.org> wrote:

> 
> ISSUE-97 (metadata content model): Poor specification of metadata content
> model [DFXP 1.0 (Editorial)]
> 
> http://www.w3.org/AudioVideo/TT/tracker/issues/97
> 
> Raised by: Sean Hayes
> On product: DFXP 1.0 (Editorial)
> 
> The current content model for <metadata> is
> Content: {any element not in TT namespace}*
> 
> This used to work, as the only non TT namespace element we had were in the
> ttm: namespace. But we now have captured the ttp:profile element as an allowed
> (syntatical) element here, but this has no semantic meaning in metadata.
> 
> In order to avoid such inadvertant captures in the future, perhaps we should
> restrict our use of 'not' in these models to only content outside any tt
> namespace, and be explicit in our inclusion of tt namespaces.
> 
> Content: ({any element in any TTM namespace} | {any element not in any TT
> namespace})*
> 
> 
> 
>

Received on Saturday, 16 May 2009 03:13:42 UTC