SML WG Teleconference Dec 11, 2008 -- 11 Dec 2008

Approval of Minutes

<johnarwe_> minutes of 12/4 http://lists.w3.org/Archives/Public/public-sml/2008Dec/0027.html

Minutes approved by unanimous consent.

Communication from CG

Nothing new.

Action Items

<lencharest> See public/2007/xml/sml/tests

Action 208: Create directories in CVS

Provided more information to Len on this action.

Action 190: Draft test case for lax fallback

MSM: No progress

Next steps on Documents

Ginny: Reports that she cannot move tests from Interoperability to Extensibility

John: Ginny can locate extensibility document from the minutes of F2F.
... Issue for WG--should we remove three features currently in Interoperability that are not Interop features or keep them?

<ginny_> From 11/20 minutes:

<ginny_> 1. Move "noSchemaNV" from "interop" list to "extensibility" list.

<ginny_> 2. Move "baseUriCheck" from "interop" list to "extensibility" list.1-

<ginny_> 3. Remove "baseUriReq" from the feature list.

<ginny_> I propose to simply remove these 3 items from the test spreadsheet since this spreadsheet contains only tests for interop features.

RESOLUTION: Proposal passes without objection.

Working Group Document

Len: Reports on progress regarding loading EPR Ref Scheme and XLink Ref Scheme.
... HMTL did not eliminate Word edit changes.

Kirk: Will work on EPR Ref Scheme references over XMas holidays.

MSM: W3C standard is to publish documents in HTML. How WGs develop such documents is up to the WG.

Len: Will prepare XLink Ref Scheme for publication.
... Will delete HTML documents from CVS and we will continue to work on the Word documents.

<ginny> http://dev.w3.org/cvsweb/2007/xml/sml/

<ginny> This is the web interface to CVS

<johnarwe_> pubrules: http://www.w3.org/2005/07/pubrules

<johnarwe_> (previous url can be found at http://www.w3.org/TR/ at the top, under "related")

<johnarwe_> http://www.w3.org/2008/09/sml-schema.xsd ..

<johnarwe_> previous is the url of a schema file someone found via a google search - have been unable to find out via webreq where it came from, lifecycle, etc

<scribe> ACTION: Michael to investigate source of the url pointing to sml-schema.xsd [recorded in http://www.w3.org/2008/12/11-sml-minutes.html#action02]

<trackbot> Created ACTION-209 - Investigate source of the url pointing to sml-schema.xsd [on Michael Sperberg-McQueen - due 2008-12-18].

New Work:

John: No new work.

Existing Work

John: Nothing productive we can do with existing work.

Test Case Discussion

John: Advises the WG to let the WG know if you are going to author a particular new test case from the wg list, so we avoid duplicate parallel effort.

Concrete responses from COSMOS to our comments.

<johnarwe_> xlsx http://lists.w3.org/Archives/Public/public-sml/2008Nov/0053.html

Rows 16 & 17

<johnarwe_> 16 - base64dtdentity.xml (decoded) contains

<johnarwe_> <?xml version="1.0"?>

<johnarwe_> <!DOCTYPE BOOK [

<johnarwe_> <!ELEMENT BOOK (P*)>

<johnarwe_> <!ELEMENT P (#PCDATA)>

<johnarwe_> <!ENTITY author "HL">

<johnarwe_> <!ENTITY copyright "IBM">

<johnarwe_> ]>

<johnarwe_> <BOOK>

<johnarwe_> chapter 1 - Intro

<johnarwe_> chapter 2 - Conclusion

<johnarwe_> &author;&copyright;

<johnarwe_> Index

<johnarwe_> </BOOK>

John: Sandy's response relates to the the difference between schema and schema documents. The absence of schema documents does not imply the absence of a schema. See SML-IF 5.4.3:

<johnarwe_> smlif says: "...the SML-IF consumer MUST compose a schema using all schema documents included in the SML-IF document and MUST use this schema to validate all instance documents in the interchange model."

<johnarwe_> thus Sandy said that a schema could be constructed, and in this case it would contain only the built-in schema components. ..as a consequence, our comments are correct.

<johnarwe_> smlif: ... 5.4.3 is read.

MSM: Agrees with Sandy interpretation of 5.4.3; therefore MSM agrees with our comments.

No objections to this interpretation.

Row 17: Base64DTDInvalidXML

<johnarwe_> decoded content: <?xml version="1.0"?>

<johnarwe_> <!DOCTYPE BOOK [

<johnarwe_> <!ELEMENT BOOK1 (P*)>

<johnarwe_> <!ELEMENT P (#PCDATA)>

<johnarwe_> <!ENTITY author "HL">

<johnarwe_> <!ENTITY copyright "IBM">

<johnarwe_> ]>

<johnarwe_> <BOOK>

<johnarwe_> chapter 1 - Intro

<johnarwe_> chapter 2 - Conclusion

<johnarwe_> &author;&copyright;

<johnarwe_> Index

<johnarwe_> </BOOK>

John: Doc invalid because Element is Book1, instance is <Book>.

MSM: Can't be schema valid, unless there is a schema. There needs to be schema.

Rows 19-23 Group: Base64EncodedDoc

John: COSMOS has updated test cases.
... WG should review these test cases for next week to determine if they are now valid.
... John will attempt to get Sandy to review these.

Row 31 uses multiple reference scheme: thus, this is not an area where we can test interop.

MSM: Implementations should tolerate another scheme even if they don't use it.
... Following discussion, MSM agrees with COSMOS response conveyed by John that we can't get to this kind of test.

John: COSMOS uses two sml reference scheme instances but must reparse the document with their own local sml uri ref scheme definition that allows 2 instances to test their consistency checking code.
... We should drop this test case as not testable under interoperability, since SML only defines a single sml ref scheme.

No objection to this approach.

Row 32: InvalidConstraintsSubstitution

John: Sandy is reviewing. 32 is correct in its new draft form; 44 still needs work.

Row 49: InValidInCompleteModel.

John: the test was written before we changed sml spec to specify how to unresolved locators; test case results (expected output) updated in COSMOS. WG should review.

Row 54:TestDocumentLocator

John: COSMOS said our comment, regarding sml:refType, is not applicable. Nothing else appears to be wrong. They think we should withdraw the comment.

Rows 55 - 56 TestRulesWithMultiple/UnderOnePattern

John: Fixed and integrated by COSMOS

Row 61: InvalidBareNameUnresolved

John: Definition of "valid"--according to our definition of valid, this is "Valid".
... from SML-perspective: this is valid

Rows 62, 66, 67, 83, 90: Same issue as row 31.

John: Multiple schemes, we don't have interop for that per earlier discussion today (above).

Row 71: SingleRef

John: Another case of definition-of-valid difference, see row 61. This is not a problem in the test case.

Rows 83 - 90: Multiple schemes issue

Row 92: ValidNilRefPSVI

John: Fixed by COSMOS bug fix.

Row 114: ValidSchematronRule

John: COSMOS not sure what the problem is. COSMOS believes this is a valid test case. We need to provide explanation if we disagree.
... Do we want to take the approach that we will assume their fixes did indeed fix the problem until someone says otherwise, or should we have someone review the modified test cases?

Ginny: Let Microsoft run the test cases and MS should inform us if something is wrong.

RESOLUTION: No objections to accepting COSMOS fixes until something is discovered to be incorrect.

Success Criteria.

John: Are we in sufficient position to discuss this topic?

Ginny and Len: We want to wait until Kumar comes back to discuss specific criteria.

John: In the meantime, I will review completeness of test cases and to code the test cases. Others should do this as well.

No other business.

Adjourn: 3:46 ET

Summary of Action Items

[NEW] ACTION: Michael to investigate source of the url pointing to sml-schema.xsd [recorded in http://www.w3.org/2008/12/11-sml-minutes.html#action02]

[End of minutes]

Updated Scribe List

Last Scribe Date        Member Name         Regrets pending 
2008-05-22              Lynn, James         Until further notice 
2008-10-29              Pandit, Kumar        12/18, 1/8 
2008-11-13              Smith, Virginia      12/18 
2008-11-20              Gao, Sandy           12/18 
2008-12-04              Charest, Len         
2008-12-11              Wilson, Kirk         
Exempt                  Arwe, John                       
Exempt                  MSM

SML WG Teleconference Dec 11, 2008

11 Dec 2008

Attendees

Contents