Re: Requirements Document Updated

At 08:32 AM 2/12/2006 -0800, Alex Milowski wrote:
>I don't think anyone is thinking that we restrict steps in a pipeline from 
>storing/producing
>non-XML outputs.  All most are saying is that "stdout" is an XML infoset.
>For example, an XSL FO processor [...] might be:

>   * a single XML element that references the document (e.g. the XHTML 
> link element):
>     <link rel="alternate" type="application/pdf" href="output.pdf"/>
>
>  * a status code:
>     <result status="ok"/>
>
>  * an Atom entry describing the new resource:
>   <atom:entry xmlns:atom="http://www.w3.org/2005/Atom">
>      <atom:title>Output Document</atom:title>
>      <atom:content type="application/pdf" src="output.pdf"/>     ...
>   </atom:entry>
>
>  * an Atom entry containing the new resource:
>   <atom:entry xmlns:atom="http://www.w3.org/2005/Atom">
>      <atom:title>Output Document</atom:title>
>      <atom:content type="application/pdf">
>      huge base64 string here
>      </atom:content>     ...
>   </atom:entry>
>
>So, all we are saying is that the "main" output is restricted to an XML 
>infoset.

OK, thanks for a great answer. A few follow up questions... Could an output
take the form of an XML document with an entity declaration in the subset
and entity reference in the XML document body? Or similarly, a <link to such
a representation?

Or are these questions that need to be asked later?

Regards,

Murray