- From: John Cowan <cowan@mercury.ccil.org>
- Date: Wed, 5 Sep 2012 15:14:57 -0400
- To: Stephen D Green <stephengreenubl@gmail.com>
- Cc: public-microxml@w3.org
Stephen D Green scripsit:
> OK, but it does make me wonder what should happen when a MicroXML
> parser tries to read an XML document that contains prefixes. Surely
> if we get that right, it will at least be able to parse it, even if
> it cannot resolve the prefix or distinguish a prefixed name from a
> non-prefixed name.
The trouble is that if we allow MicroXML parsers to accept QNames
without knowing the XML namespace rules, we are licensing MicroXML
authors/generators to create QNames without knowing the namespace rules.
And while that is technically compliant with XML 1.0, it is an insuperable
barrier to injecting such documents into an XML toolchain with more than
just a parser into it.
So the current consensus is to ban colons, which means there is no way
to break the namespace rules. Allowing just xml:* names would not have
broken them either, but didn't get enough support.
--
John Cowan cowan@ccil.org http://ccil.org/~cowan
In computer science, we stand on each other's feet.
--Brian K. Reid
Received on Wednesday, 5 September 2012 19:15:19 UTC