Re: xml:* attributes

On Thu, Aug 16, 2012 at 2:54 AM, David Carlisle <davidc@nag.co.uk> wrote:

> On 15/08/2012 22:13, John Cowan wrote:
>
>> On reflection, I'd like to see option A-prime, in which colons are
>> not allowed except in the form "xml:something".  This has no
>> namespace issues and preserves compatibility with XML 1.0 NS, but
>> allows the standard attributes to still work.
>
>
> I don't think allowing xml: is a very big (or particularly problematic)
> extension to the _markup language_ compared to "no colons" but I would
> still argue against it as it only makes sense to do this if we
> complicate the data model to an extent that I'm not sure makes sense for
> micro-xml.
>

James's proposal of Option C would affect the data model only to
grandfather in xml:* atributes defined in the spec, and yes that might be
too much.  John's very similar proposal of A-prime omits any even virtual
effect on the data model, so there is no complication.  What it does is
leave any interpretation that could affect the model to a separate layer,
such as...



> If we allow xml:id then it should be an ID. That is, it should have some
> uniqueness properties and the value should work as a fragment identifier
> if micro-xml documents are served with a suitable mime type.
>
(Your Editor's draft just reserves xml:id but explicitly makes its ID
> properties out of scope. I am not sure that really helps
> interoperability or compatibility with XML.)


The xml:id spec already covers this, and the separate statement of
reservation of "xml..." names being formulated in the other thread would
guard against abuse of standard xml:* attributes.


-- 
Uche Ogbuji                       http://uche.ogbuji.net
Founding Partner, Zepheira        http://zepheira.com
http://wearekin.org
http://www.thenervousbreakdown.com/author/uogbuji/
http://copia.ogbuji.net
http://www.linkedin.com/in/ucheogbuji
http://twitter.com/uogbuji