RE: [xml-dev] "Introducing MicroXML, Part 1: Explore the basic principle

[Pete Cordell <petexmldev@codalogic.com>]

On 1 July 2012 16:01, David Lee <dlee@calldei.com> wrote:
> John Cowan Sez ...
>> This all seems to me nothing more than a vast to-do about whether
>> a general-purpose href attribute ought to be xlink:href or xml:ref.
>> I cannot take the question seriously.  "Parturient montes, nascetur
>> ridiculus mus."
>
> My take is a little more serious.  If an attribute is part of the xml 
> namespace then there is a presumption that all consumers of "XML" 
> understand and apply the semantics.  I would think it would be a 
> "must do".  But putting an attribute in another namespace makes it 
> a "do if you want to support that thingy".   I think this is a big 
> difference.

I would take the completely opposite view.  Xlink and friends are there for vocabulary designers to use, but they imply no requirement for support by a basic XML parser and a vocabulary designer is at liberty to define their own set of attributes that do the same thing or something slightly different.

I see no reason why some set of attributes in the XML namespace should not be available to vocabulary designers on the same basis.

Then, in much the same way that xmlns in MicroXML is an application level concept, xml:href could also be a purely _optional_to_understand_ application level concept.

Thus it seems to me that the XML namespace should be as big as it needs to be but no bigger.  But really it doesn't matter whether it includes the kitchen sink because only applications that are interested in those features (on a pick and mix basis) will be burdened by them.

The benefit of course is that documents won't have to include xmlns:xlink="http://www.w3.org/1999/xlink"; and developers don't have to deal with understanding namespaces, which is one of the main motivations of MicroXML.  If we require the use of xmlns:xlink then all we've done is move namespaces up to the application level, which may make us XML fanatics more satisfied with our architecture, but just foists the problem directly onto developers and does nothing to alleviate the general confusion they have about using namespaces.

(P.S. Sorry for my late entry into this debate.  I was waiting for Michael Kay's input. But since he's abstained I now have to work things out for myself!)

Pete Cordell
Codalogic Ltd
Interface XML to C++ the easy way using XML C++
data binding to convert XSD schemas to C++ classes.
Visit http://codalogic.com/lmx/ or http://www.xml2cpp.com
for more info