SML WG Conference Call

16 Oct 2008

See also: IRC log


Pratul, John, Kumar, Kirk
Sandy, MSM, Ginny
Pratul Dublish
Kirk Wilson




<pratul> Agenda is at http://lists.w3.org/Archives/Public/public-sml/2008Oct/0008.html

<scribe> scribenick: kirkw

<scribe> scribe: Kirk Wilson

<johnarwe_> minutes at http://lists.w3.org/Archives/Public/public-sml/2008Oct/att-0006/20081002-sml-minutes.html

Approval of minutes from 10/2

RESOLUTION: Minutes approved without objection.

Review Schema working group's responses to LC1 comments

Issue 5053: http://www.w3.org/Bugs/Public/show_bug.cgi?id=5153

John: Issue has to do with word order and clarification.

Pratul: Issue is, Shall we endorse the resolution?

RESOLUTION: WG endorses the resolution.

Issue 5155: http://www.w3.org/Bugs/Public/show_bug.cgi?id=5155

RESOLUTION: WG endorses the resolution.

Action Items

John: Only two open, from MSM and Pratul for draft of XLink note.

Pratul: Will have XLink note for F2F. MSM not here today.

Latest draft of Test Case Document

Discussion of section 2 of test doc current draft under discussion

Correction to p. 1 line 25:

<johnarwe_> inconsistency betw 1.25 and 3.21-22 to be corrected

<johnarwe_> btw, for the IRC record, for today I am repping IBM since Sandy is not able to attend

<Kumar> from : Therefore, each test will be represented by an SML-IF document.

<Kumar> to : Therefore, all tests, except the tests that test the locator element, will be represented by an SML-IF document.

<johnarwe_> 2.16 documentS

RESOLUTION: Text as pasted in IRC is approved.

<johnarwe_> 4. 16 resultS

<johnarwe_> 4.16 and -> or

<johnarwe_> 4.17 resultS

<johnarwe_> 4.23 This -> Comparing test results (so it refers back to 1st sentence, not 2nd, which seems like the original intent)

John: bottom of p. 4.37: We have additional question if SML-IF document is valid, whether the model is SML valid. This leads to the possibility of a tertiary value of the results. Results, therefore, are not simply a boolean value.
... There are three states: SML-IF invalid vs. SML-IF valid (which can be SML valid or invalid)

Kumar: Addressed by lines 1 - 8 on p. 5.
... This is not a problem for the two implementations that we know. It will be clear from the test time what the source of the error is.

John: Boolean is correct: Issue is what can be guaranteed from the spec and what you can know as a human. The two are not the same.

RESOLUTION: No objections from current attendees to approving the test-plan doc with the specific change on p. 1.

Review of COSMOS Test Plan.

See Ginny's email.

<pratul> http://lists.w3.org/Archives/Member/member-sml/2008Oct/0010.html

<pratul> an SML reference (sml:ref = true) using only unrecognized schemes. (#1 above)

Discussion: SML references using unrecognized schemes.
... What is the expected result of the test?

John: If targetRequired, then SML reference is invalid.

Kumar: Doesn't see much value in writing such a test case. If both implementations doesn't understand the reference schemes, there is no issue of interoperability.

John: We need to answer the question of whether we are starting with COSMOS and then just discuss additional test cases?
... Pratul agrees we should start with this question.

RESOLUTION: Agreed without objection to start with accepting the COSMOS test suite.

Returning to considering Ginny's list:

<pratul> - an SML reference (sml:ref = true) using only unrecognized schemes. (#1 above)

Kumar: Proposal is NOT to add it.

John: If Ginny (or anyone else) was to write such a case, I would not reject it.

RESOLUTION: The group will not write such a case, but if anyone were to write a case, we would accept it.

Second Test Case Issue: does not look like there are tests that test the necessary processing to identify identical targets (section 4.2.3) E.g., bullet #2 is not tested.

Kumar: MS implementation could not test such a condition, since it supports only the SML URI reference scheme.

Pratul: We have different aliases pointing to the same element.

<johnarwe_> sml 4.2.3 #2 starts Otherwise, a model validator MUST consider both targets to be different when

Pratul: Proposal is to add this test case.

RESOLUTION: We should add a test case to cover this scenario.
...Pratul: have one or more test cases.

Third Test Case Issue: I don't see deref() tests for each bullet in section 4.2.7, 1.b.

Pratul: Proposal is to add test case to cover this scenario: 1.b test case.

<pratul> Proposal: Add test case(s) to cover 4.2.7, 1(b)

Kumar: Since MS supports only one scheme, MS could not test it.

RESOLUTION: If anybody is willing to write the test case, we will accept it.
... Attendees are "neutral" to this test case.

Fourth Test Case Issue: no test for section 4.3.1, bullet 1 (wrong namespace for 'uri') and bullet 1.a.

<pratul> Ginny: no test for section 4.3.1, bullet 1 (wrong namespace for 'uri') and bullet 1.a.

<Kumar> Third bullet case will fall into the optional features test bucket.

Pratul: Proposal is to add test cases for this scenario.

RESOLUTION: We should add a test case to cover this scenario.
...Kumar: there may be a test case for this.

Fifth Test Case Issue: no targetRequired tests for derivation by restriction or substitution groups (there are tests for these in targetElement and targetType) - section, bullet 1.b and section (for targetRequired).

Pratul: Proposal is to add these test cases.

Kumar: Agreed.

RESOLUTION: We should add test cases to cover this scenario.

Sixth Test Case Issue: no deref() test for sml:selector or sml:field, sections - bullets 1 and 2.

Kumar: We have test cases for this; also COSMOS.

<Kumar> InValidKeyDuplicate.xml

Kumar: Ginny may mean what happens if there are invalid XPath.

Pratul: We need more information from Ginny regarding what she means and go on from there.
... Pratul will write Ginny an email after the call.

Seventh Test Case Issue: no test for section, bullet 4.

Pratul: Proposal is to add test cases for this, if COSMOS has no test cases for this scenario.

RESOLUTION: Agreed, no objections.

Eighth Test Case Issue: acyclic tests do not mention "intra-document references" so I assume there may not be a test for this. The tests only mention "inter-document references".

Pratul: Proposal is to add test cases to cover intra-document acyclic constraint for intra-document references.

RESOLUTION: WG agrees with no objections.

Pratul: We will meet next week.

Meeting adjourned: 3:38 ET

Summary of Action Items

[End of minutes]

Updated Scribe List
Last Scribe Date        Member Name         Regrets pending 
2008-05-22              Lynn, James         Until further notice
2008-07-10              McCarthy, Julia     Until further notice
2008-09-18              Smith, Virginia   
2008-09-25              Kumar, Pandit
2008-10-02              Gao, Sandy
2008-10-16              Wilson, Kirk
Exempt                  Arwe, John           
Exempt                  Dublish, Pratul
Exempt                  MSM