Re: [xml-dev] RFC for XML Object Parsing

[Liam R E Quin <liam@w3.org>]

On Mon, 2014-03-24 at 20:26 -0600, Brian Aberle wrote:
> [...] If the world was to increment the XML spec

Unlikely, I'm afraid...

>  it would seem logical that attribute order would be preserved

There's a theoretical reason not to - they're properties of the element,
logically. Appendix F of the ISO SGML standard compared attributes to
command parameters.

Yes, it's one of a number of places where XML could be made easier to
implement, but also a reinder that XML (like SGML) is about modelling
documents rather than defining a format convenient for any specific
program to parse. It's equally inconvenient for all programs :)

On the other hand, being able to say, e.g., that a namespace declaration
must come earlier in a start tag than any use of that namespace (other
than the element name itself which for historical reasons is always
first) would for sure facilitate processing.

Round-trip parsing of JSON similarly suffers from reordering of hash
values.

For my part I'd like to have attributes with structure, e.g. by
designating an xml:attributes element as the first non-whitespace child
of an element, but the result you get isn't really XML, and it quickly
gets inconvenient for hand markup of documents.


-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org freenode/#xml