Re: Additional 'namespace' attribute on content element?

2007-01-05 Thread Martin Duerst

At 16:13 07/01/05, Asbj\x8F\xD3n Ulsberg wrote:

On Thu, 04 Jan 2007 17:00:29 +0100, Jan Algermissen  
[EMAIL PROTECTED] wrote:

 on the NewsML list, an issue came up that due to they lack of a MIME  
 type for NewsML using NewsML as Atom content is somewhat problematic[1];  
 I think this is the case with most of the more interesting XML  
 applications out there.

The more interesting XML applications out there should get a proper MIME  
type, that's all. Nothing wrong with Atom in this case, imo.

Sorry, wrong.

Using +xml at the end of a Mime Type is a good convention, but it
is only a convention. There is no requirement for a Mime Type that
uses XML to end in +xml. Also, the +xml convention was established
several years later than XML itself.

So I think that Atom is a bit out on it's egde if it says
'if you don't have a +xml Mime Type, you're not XML'.

Regards,Martin.




#-#-#  Martin J. Durst, Assoc. Professor, Aoyama Gakuin University
#-#-#  http://www.sw.it.aoyama.ac.jp   mailto:[EMAIL PROTECTED] 



Re: Additional 'namespace' attribute on content element?

2007-01-05 Thread Henri Sivonen


On Jan 5, 2007, at 09:47, Martin Duerst wrote:


So I think that Atom is a bit out on it's egde if it says
'if you don't have a +xml Mime Type, you're not XML'.


No, Atom says that if you don't use application/xml or an +xml type,  
you don't get to use an embedded subtree and your format gets the  
kind of treatment that non-XML formats get.


Atom is quite reasonable here. An XML vocabulary that doesn't have a  
MIME type that follows the convention *and* doesn't have a namespace  
is itself a badly-behaved XML vocabulary.


--
Henri Sivonen
[EMAIL PROTECTED]
http://hsivonen.iki.fi/




Re: Additional 'namespace' attribute on content element?

2007-01-05 Thread A. Pagaltzis

* James Aylett [EMAIL PROTECTED] [2007-01-04 18:55]:
 I don't see a big difference between dispatch on namespace of
 child of content (which is, I believe, legal in Atom 1.0) and
 dispatch on namespace attribute on content (which is what I
 think you're proposing).

Except when there is nothing you can dispatch on because the
vocabulary in question is not in a namespace.

Regards,
-- 
Aristotle Pagaltzis // http://plasmasturm.org/



Re: Additional 'namespace' attribute on content element?

2007-01-05 Thread Mark Baker


On 1/4/07, Jan Algermissen [EMAIL PROTECTED] wrote:

I recall Mark Baker using something similar in his former RDF Forms draft:

rf:Container xmlns:rf=http://www.markbaker.ca/2003/rdfforms/;
  xmlns:rdf=http://www.w3.org/1999/02/22-rdf-syntax-ns#;
  rdf:about=http://shoes.example.com/order-processor/;
  rf:acceptedMediaTypeapplication/xml/rf:acceptedMediaType
  rf:acceptedNamespace 
rdf:resource=http://shoe-standards.example.org/orders/shoes//  
  rf:intent rdf:resource=http://shoe-standards.example.org/order-shoes//
/rf:Container


That just describes an expectation ala app:accept, rather than any
inline content ala the type attribute on atom:content.

Mark.
--
Mark Baker.  Ottawa, Ontario, CANADA. http://www.markbaker.ca
Coactus; Web-inspired integration strategies  http://www.coactus.com



Re: Additional 'namespace' attribute on content element?

2007-01-05 Thread Asbjørn Ulsberg


On Fri, 05 Jan 2007 13:26:46 +0100, Henri Sivonen [EMAIL PROTECTED] wrote:

Atom is quite reasonable here. An XML vocabulary that doesn't have a  
MIME type that follows the convention *and* doesn't have a namespace is  
itself a badly-behaved XML vocabulary.


My point exactly.

--
Asbjørn Ulsberg -=|=-http://virtuelvis.com/quark/
«He's a loathsome offensive brute, yet I can't look away»



Re: Additional 'namespace' attribute on content element?

2007-01-05 Thread Jan Algermissen


Hi Mark,

On Jan 5, 2007, at 6:21 PM, Mark Baker wrote:



On 1/4/07, Jan Algermissen [EMAIL PROTECTED] wrote:
I recall Mark Baker using something similar in his former RDF  
Forms draft:


rf:Container xmlns:rf=http://www.markbaker.ca/2003/rdfforms/;
  xmlns:rdf=http://www.w3.org/1999/02/22-rdf-syntax-ns#;
  rdf:about=http://shoes.example.com/order-processor/;
  rf:acceptedMediaTypeapplication/xml/rf:acceptedMediaType
  rf:acceptedNamespace rdf:resource=http://shoe- 
standards.example.org/orders/shoes//  
  rf:intent rdf:resource=http://shoe-standards.example.org/order- 
shoes//

/rf:Container


That just describes an expectation ala app:accept, rather than any
inline content ala the type attribute on atom:content.



Sure. Sorry if you feel misrepresented. I only meant to provide an  
example of how one addressed the lack of distinguished MIME types for  
some formats by using the formats namespace instead.


Jan



Mark.
--
Mark Baker.  Ottawa, Ontario, CANADA. http://www.markbaker.ca
Coactus; Web-inspired integration strategies  http://www.coactus.com





Additional 'namespace' attribute on content element?

2007-01-04 Thread Jan Algermissen

Hi,

on the NewsML list, an issue came up that due to they lack of a MIME type for 
NewsML using NewsML as Atom content is somewhat problematic[1]; I think this is 
the case with most of the more interesting XML applications out there.

Is there any chance to extend/revise Atom to allow an attribute on the content 
element that allows to specify the namespace of XML content given the MIME type 
is declared as application/xml or text/xml?

Actually, there is no posibility to do this as an Atom extension, is there?

What is the opinion of the compund document formats experts on this?

Note:

I recall Mark Baker using something similar in his former RDF Forms draft:

rf:Container xmlns:rf=http://www.markbaker.ca/2003/rdfforms/;
  xmlns:rdf=http://www.w3.org/1999/02/22-rdf-syntax-ns#;
  rdf:about=http://shoes.example.com/order-processor/;
  rf:acceptedMediaTypeapplication/xml/rf:acceptedMediaType
  rf:acceptedNamespace 
rdf:resource=http://shoe-standards.example.org/orders/shoes//  
  rf:intent rdf:resource=http://shoe-standards.example.org/order-shoes//
/rf:Container

Jan


[1] http://tech.groups.yahoo.com/group/newsml/message/2424
[2] http://www.markbaker.ca/2003/05/RDF-Forms/rdf-forms-old1.html



Re: Additional 'namespace' attribute on content element?

2007-01-04 Thread James M Snell



Jan Algermissen wrote:
 Hi,
 
 on the NewsML list, an issue came up that due to they lack of a MIME type 
 for NewsML using NewsML as Atom content is somewhat problematic[1]; I think 
 this is the case with most of the more interesting XML applications out there.
 
 Is there any chance to extend/revise Atom to allow an attribute on the 
 content 
 element that allows to specify the namespace of XML content given the MIME 
 type 
 is declared as application/xml or text/xml?
 
 Actually, there is no posibility to do this as an Atom extension, is there?
 

Of course this can be done as an extension. The bigger problem is that
NewsML doesn't have an XML Namespace to declare.  At least, I don't see
one in version 1.2

- James



Re: Additional 'namespace' attribute on content element?

2007-01-04 Thread James Holderness


Jan Algermissen wrote:

on the NewsML list, an issue came up that due to they lack of a
MIME type for NewsML using NewsML as Atom content is somewhat
problematic[1];


From my brief reading of the thread, I got the impression that NewsML did 
have a media type (text/vnd.IPTC.NewsML) - it just isn't recognised as an 
XML media type. If that's correct, then surely you can just escape the 
content as you would with html or any other text/ media type. Obviously 
it's not ideal, but it is workable.


If you think it's vitally important that the content be embedded as 
unescaped XML, then I would think a more practical solution would be to get 
a +xml media type rather than trying to extend Atom.


Regards
James



Re: Additional 'namespace' attribute on content element?

2007-01-04 Thread James Aylett

On Thu, Jan 04, 2007 at 08:00:29AM -0800, Jan Algermissen wrote:

 Is there any chance to extend/revise Atom to allow an attribute on
 the content element that allows to specify the namespace of XML
 content given the MIME type is declared as application/xml or
 text/xml?
 
 What is the opinion of the compund document formats experts on this?

I wouldn't class myself as an expert, but why do we need an attribute?
How about something like (please not text/xml, by the way :-):

--
atom:content type=application/xml
  newsml:whatever xmlns:newsml=newsml namespace URI
...
  /newsml:whatever
/atom:content
--

I don't see a big difference between
dispatch-on-namespace-of-child-of-content (which is, I believe, legal
in Atom 1.0) and dispatch-on-namespace-attribute-on-content (which is
what I think you're proposing).

James

-- 
/--\
  James Aylett  xapian.org
  [EMAIL PROTECTED]   uncertaintydivision.org



Re: Additional 'namespace' attribute on content element?

2007-01-04 Thread Asbjørn Ulsberg


On Thu, 04 Jan 2007 17:00:29 +0100, Jan Algermissen  
[EMAIL PROTECTED] wrote:


on the NewsML list, an issue came up that due to they lack of a MIME  
type for NewsML using NewsML as Atom content is somewhat problematic[1];  
I think this is the case with most of the more interesting XML  
applications out there.


The more interesting XML applications out there should get a proper MIME  
type, that's all. Nothing wrong with Atom in this case, imo.


--
Asbjørn Ulsberg -=|=-http://virtuelvis.com/quark/
«He's a loathsome offensive brute, yet I can't look away»