Re: Considering the standalone declaration

On Mon, Oct 17, 2011 at 11:18:47AM -0400, John Cowan wrote:
> Grosso, Paul scripsit:
> 
> > Henry: The browsers aren't going to pay attention to the standalone
> > declaration.
> > ... Unless we change the XML spec to change the default. The problem is
> > that the default is standalone=no. So if we ask the browsers to change
> > to make standalone=no an error, we'll break all XHTML. It's a lose-lose
> > situation.
> 
> [etc.]
> 
> I must be missing something here, because this discussion seems to me
> entirely pointless.  The standalone="yes" declaration is a hint to a
> non-validating processor (NVP) that it does not have to read any external
> parameter entities including the external DTD subset (henceforth EPEs)
> because none of them can possibly affect the document infoset.  It has
> no meaning to a validating processor, which must read all entities in
> any case.  Likewise, it has no meaning to a NVP which never reads EPEs:
> browser XML processors fall into this category.

  I agree, I don't see a problem with the status quo.
The idea of trying to change the default in XML is IMHO a dead end,
we are moving away from external subset loading, not trying to get this
back in, especially as this part of the spec is often misunderstood.
And we definitely don't want to change the semantic of the zillion
existing documents (and make a number of them suddenly not well-formed
which is a clasic in the testing of standalone, where a wrong
standalone declaration caught up by a validating parser who loads
the external subset can lead to a not-well formed document).

Daniel

-- 
Daniel Veillard      | libxml Gnome XML XSLT toolkit  http://xmlsoft.org/
daniel@veillard.com  | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library  http://libvirt.org/

Received on Tuesday, 18 October 2011 10:25:25 UTC