RE: document node attributes

I suggest your answers illuminate the fundamental difference of opinion you and I have of "What Is XML"
While I do recall (but cant find it now) that an early tagline for XML was "For the Web", I argue that is no longer the case.
That XML has surpassed the web in so many ways that adding things to the core of XML specifically for the Web is a detriment to core XML.
Does the message that a temperature controller in a car send to the engine controller need hyperlinking ?
I dont joke.  XML is used in so many non-web places that I would *guess* that the amount of non-web use of XML is atleast 1millionx more than web oriented  XML,
but again that is a guess and impossible to validate.
But I can state with high confidence the web is only a small part of what XML is for.  This my main objection to insisting that hypermedia does not belong in the xml:* namespace.  My other objection is practical.  I have complete confidence the XML WG views the issue the same way - they have stated it on multiple occasions,
so for practical reasons I am not thrilled to be working on a project that is doomed from the start.

Given that, and given your arguments are 100% based on XML for use on the Web ... I suggest

Maybe a possible road to "success" of hypermedia is defining a "Profile" for XML designed for the web. 
Much like MicroXML which is designed as a subset of XML intended for simpler applications, one could imagine a simpler concept designed for "XML On the web"
Then we are free to write our own rules about what is part of the core spec or not



-----------------------------------------------------------------------------
David Lee
Lead Engineer
MarkLogic Corporation
dlee@marklogic.com
Phone: +1 812-482-5224
Cell:  +1 812-630-7622
www.marklogic.com



-----Original Message-----
From: Rushforth, Peter [mailto:Peter.Rushforth@NRCan-RNCan.gc.ca] 
Sent: Monday, June 24, 2013 7:56 AM
To: David Lee; liam@w3.org
Cc: public-xmlhypermedia@w3.org
Subject: RE: document node attributes

Hey David,

> 
> 1) What percentage of XML documents to you believe are "on 
> the internet"  and processed by tools which have 
> "accessibility to the internet"

I have no idea.   I'm focused on the Web, in contrast to the Internet.
The Web has apparently very little XML on it, at least relatively speaking.
I suppose that percentage is shrinking all the time, too.

Which is a bit odd, when you think of the push for "linked data".
Certainly HTML isn't "meant" to be a data encoding.

> 1A) what percentage of XML Documents even have a base uri ? 

Everything on the Web has a URI, or an address from which it
is obtained. But I'm dodging the question, I know.  It's a facility
that is very useful.  I use it.  It can dramatically reduce file
size where URIs form a lot of the content. 

> 
> 2) In your opinion what percentage of expected usage should 
> be a requirement in order to add something to the xml: namespace ?

I doubt that is how the xml: namespace threshold is established.

But if it is useful, enhances interoperability etc.,
why wouldn't it get a place there?


> 
> 3) In your opinion, does adding a attribute to the xml: 
> namespace require that  processors recognize it as valid ?

That's a good question.  I think there are things that an xml
processor could do for a client layer, for example to provide
overridable default behaviour for @href and @src i.e. fetch the
link with the suggested media type preference.  


> 
> 4) In your opinion what makes hypermedia extensions "more 
> central" to xml than any other extensions like XLink or 
> Dublin Core or XSD  ...

The Web needs a _simple_ reliable standard in order to scale.


> 
 , could you argue that hypermedia is *more 
> exaulted* or *more special* or *more central* to XML then 
> other technologies ? such that it deserves special treatment 
> more than any other technology by the entire world ?

Yes, I think the Web is a special case.  As mentioned, the
number one goal of XML is success on the internet.  You could
say, (I do say), that if you fail on the Web, you fail on the Internet.

The World Wide Web Consortium is named after the Web.

> 4A) Eg. If something as basic to XML as schema, which is in 
> its own namespace *(e.g. 
> xsi:schemaLocation="http://NamespaceTest.com/Purchase 
> Main.xsd" )

I've had this discussion before with David C. on this list.   xsi:schemaLocation 
is itself a form of hypertext reference(s), albeit with added semantics.

I can't claim that XML Schema would have used xml:href and xml:src
had they been available anyway, but I believe there are other
less complicated applications that could and would be in a position
to use xml:href and xml:src etc.  Simplicity is the key to scale.

Cheers,
Peter

Received on Friday, 28 June 2013 11:54:37 UTC