The current content has been slightly changed since the bug has been opened.
 
Current content for section 4.2.2 :

4.2.2 EPR Scheme
The EPR reference scheme MUST be implemented by using an instance of the wsa:EndpointReference global element declaration [WS-Addressing Core] as a  child element of the SML reference element. Instances of the EPR reference scheme MUST NOT be interpreted as inter-document references in the context of an SML-IF document.

The following example illustrates how the EnrolledCourse reference that references course PHY101 in MIT university can be represented using the EPR scheme:

[Example Under Construction]

<EnrolledCourse xmlns="http://www.university.example.org/ns" sml:ref="true">
  <wsa:EndpointReference
       xmlns:u="http://www.university.example.org/schema">
    <wsa:Address>http://www.university.example.org</wsa:Address>
  </wsa:EndpointReference>
</EnrolledCourse>


Changed to :

The EPR reference scheme MUST be implemented by using an instance of the wsa:EndpointReference global element declaration [WS-Addressing Core] as a  child element of the SML reference element. 

The following example illustrates how the EnrolledCourse reference that references course PHY101 in MIT university can be represented using the EPR scheme:

<EnrolledCourse xmlns="http://www.university.example.org/ns" sml:ref="true">
  <wsa:EndpointReference
       xmlns:u="http://www.university.example.org/schema">
    <wsa:Address>http://www.university.example.org</wsa:Address>
  </wsa:EndpointReference>
</EnrolledCourse>

What changed :
1.Removed : Instances of the EPR reference scheme MUST NOT be interpreted as inter-document references in the context of an SML-IF document. ( this is already in IF; SML should not refer to IF )
2. Removed: [Example Under Construction] ( you proposed to change it with Valid example but I feel this is redundant with the paragraph above )
3. Replaced : 
....as a  child element of the SML reference element.
with ...as a  child of the SML reference element.

reopening.. 

I meant to mark it for review; changed keyword to needsReview

It is not at all clear as to how this example

<EnrolledCourse xmlns="http://www.university.example.org/ns" sml:ref="true">
  <wsa:EndpointReference
       xmlns:u="http://www.university.example.org/schema">
    <wsa:Address>http://www.university.example.org</wsa:Address>
  </wsa:EndpointReference>
</EnrolledCourse>

references the course PHY101 in MIT university. Can we make the <wsa:Address> element more specific - say www.phy101.mit.university.org?  

In reply to comment #3:

The actual content of the sample is not intended to be addressed by this defect.
Defect 4637, dealing with EPR scheme will take care of the sample's content

Regarding the following content from comment #1 (near the end):
"What changed :
1.Removed : Instances of the EPR reference scheme MUST NOT be interpreted as
inter-document references in the context of an SML-IF document. ( this is
already in IF; SML should not refer to IF )
"

This statement should not have been removed. The resolution for bug #4819 (See comment #2 in bug #4819) states that this statement belongs in SML *not* in SML-IF.

re comment #5

Section 4.2.2 EPR Schema, SML core, already has this sentence which addresses your concern:

3. The EPR Scheme can not be used as an interdocument reference when used in an SML-IF [SML-IF 1.1] document.

Comment on the sentence above : this is normative so we shoud probably reword this as :

3. The EPR Scheme >>MUST NOT<< be used as an interdocument reference when used in an SML-IF [SML-IF 1.1] document.

The normative statements in comment #5 seem to be inconsistent with the proposal made in 5119, which is proposed for SML-IF:

(5) to the paragraph starting "In contrast, the wsa:address..." append
"A sufficiently constraining reference scheme definition, e.g. one that
determined the protocol binding used to interact with the endpoint, MAY assert
that instances of the reference scheme are inter-document references."

The case where it is possible for an EPR to be inter-document reference occurs when the protocol binding is sufficient constrained by the reference scheme defintion.  I don't believe we discussed how that is to be done, so I suspect it is difficult at this point to say anything about the normative status of the EPR reference being an inter-document reference.

Since this defect was mainly intended to deal with minor editorial corrections on section 4.2.2 ( see John's initial description ) and NOT with the validity of the sample or EPR support ( which are to be addressed under 5242 and 4637 ) I propose to close this defect as fixed and move the discussion from comments #3 and #7 to that thread.

Dealing in one defect with more than it was initially intended by the defect originator is confusing and hard to manage from an editorial perspective.

Full text proposal for section 4.2.2 epr scheme.  This text also resolves issues 4637 and 5242

http://lists.w3.org/Archives/Public/public-sml/2007Nov/0250.html

Changed back keyword from hasProposal to needsReview ( see comment #8 on why I did that)

Resolving "won't fix" since the EPR related text has been removed as a part of fix to bug# 4637.

("won't fix" is the not the most correct resolution since part of the fix was already applied by Valentina. Though this part got removed due to 4637 fix).