<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
<!DOCTYPE bugzilla SYSTEM "https://www.w3.org/Bugs/Public/page.cgi?id=bugzilla.dtd">

<bugzilla version="5.0.4"
          urlbase="https://www.w3.org/Bugs/Public/"
          
          maintainer="sysbot+bugzilla@w3.org"
>

    <bug>
          <bug_id>4682</bug_id>
          
          <creation_ts>2007-06-21 15:59:30 +0000</creation_ts>
          <short_desc>Attribute based reference schemes</short_desc>
          <delta_ts>2007-09-19 23:21:03 +0000</delta_ts>
          <reporter_accessible>1</reporter_accessible>
          <cclist_accessible>1</cclist_accessible>
          <classification_id>1</classification_id>
          <classification>Unclassified</classification>
          <product>SML</product>
          <component>Core</component>
          <version>unspecified</version>
          <rep_platform>PC</rep_platform>
          <op_sys>Windows XP</op_sys>
          <bug_status>RESOLVED</bug_status>
          <resolution>FIXED</resolution>
          
          
          <bug_file_loc></bug_file_loc>
          <status_whiteboard></status_whiteboard>
          <keywords></keywords>
          <priority>P2</priority>
          <bug_severity>normal</bug_severity>
          <target_milestone>Second draft</target_milestone>
          
          
          <everconfirmed>1</everconfirmed>
          <reporter name="Sandy Gao">sandygao</reporter>
          <assigned_to name="Kumar Pandit">kumarp</assigned_to>
          
          
          <qa_contact name="SML Working Group discussion list">public-sml</qa_contact>

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>15496</commentid>
    <comment_count>0</comment_count>
    <who name="Sandy Gao">sandygao</who>
    <bug_when>2007-06-21 15:59:30 +0000</bug_when>
    <thetext>Should SML allow attribute based reference schemes?

SML currently uses xsi:nil=&quot;true&quot; as an indication for &quot;this reference has no value&quot;. In XML Schema, xsi:nil only controls whether an element has any child content, and does *not* control whether the element carries attributes or not.

As a consequence, if a reference schema uses attributes, e.g.

&lt;my:myRefElement sml:ref=&quot;true&quot; my:reference=&quot;location&quot; xsi:nil=&quot;true&quot;/&gt;

then the xsi:nil based approach to determine whether a reference is nil is not reliable. We need to decide whether we want to support attribute based schemes. If we do, we can&apos;t rely on xsi:nil for nil references.
whether we should require reference pointers be children.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>16122</commentid>
    <comment_count>1</comment_count>
    <who name="Kumar Pandit">kumarp</who>
    <bug_when>2007-08-10 00:20:10 +0000</bug_when>
    <thetext>Refer to the following related bug:
4884 update the definition of a null reference to include the
unimplemented-schemes case

If we update the definition of a null ref as described in the bug, we no longer need to rely on xsi:nil value to define null ref. Consequently, the issue mentioned in this bug goes away.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>16417</commentid>
    <comment_count>2</comment_count>
    <who name="Pratul Dublish">pratul.dublish</who>
    <bug_when>2007-09-07 03:52:12 +0000</bug_when>
    <thetext>I think that the bug description is a bit confusing since it mixes the issues of supporting attribute-based reference schemes and identification of null references. The introduction of sml:nilref solves the latter issue, but the issue of attribute-based schemes still remains. </thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>16531</commentid>
    <comment_count>3</comment_count>
    <who name="John Arwe">johnarwe</who>
    <bug_when>2007-09-13 18:42:14 +0000</bug_when>
    <thetext>Consensus 2007-09-13 telecon: 4884 handles null ref portion of this bugzilla.

Consensus 2007-09-13 telecon: attribute based ref schemes are allowed in SML.  If we find anything in the spec that doesn&apos;t work well with attribute-based reference schemes, it is grounds for raising an issue.
</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>16552</commentid>
    <comment_count>4</comment_count>
    <who name="Kumar Pandit">kumarp</who>
    <bug_when>2007-09-14 06:43:16 +0000</bug_when>
    <thetext>added text:

A reference scheme MAY use elements, attributes or both to capture the information necessary to identify a resource. 

</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>16693</commentid>
    <comment_count>5</comment_count>
    <who name="Sandy Gao">sandygao</who>
    <bug_when>2007-09-19 17:34:38 +0000</bug_when>
    <thetext>The change mentioned in comment #4 answer the immediate question about schemes that use attributes, but the new text is assuming that schemes will always uses some elements or attribute, which as far as I know has not been adopted as a requirement for defining a scheme.

One example we discussed before is a scheme that always resolves to the root element of the current document, where it doesn&apos;t depend on either elements or attributes.

And a scheme may want to define its behavior in terms of a processing instruction, which would seem to be a perfectly valid scheme.

Suggest to use something similar to:

&quot;A reference scheme normally uses, but is not required to use, child elements, attributes or both to capture ...&quot;</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>16698</commentid>
    <comment_count>6</comment_count>
    <who name="Kumar Pandit">kumarp</who>
    <bug_when>2007-09-19 23:20:19 +0000</bug_when>
    <thetext>changed the text as suggested by Sandy:

A reference scheme normally uses, but is not required to use, child elements, attributes or both to capture the information necessary to identify a resource. 
</thetext>
  </long_desc>
      
      

    </bug>

</bugzilla>