TAG TC 2010-05-11 Meeting Minutes

["Dennis E. Hamilton" <dennis.hamilton@acm.org>]

Test Assertions Guidelines (TAG) TC Teleconference Meeting


Logistics:
----------

Tuesday 2pm (PDT), 21:00Z, 11 May 2010

Host confcall: Fujitsu
US Toll Free: 877-995-3314
US Toll/International: 210-339-1806
Passcode: 9589308

Attendance and Quorum
---------------------

MEMBERS IN ATTENDANCE

5/5 Voting Members attending: 3/5 requirement for quorum was satisfied

 * Jacques Durand, Chair (Fujitsu)
 * Dennis Hamilton, Secretary (Individual) 
 * Frederick Boland (NIST)
 * Kevin Looney (Oracle)
 * Paul Rank (Oracle)

 [*: Voting member at start and conclusion of the meeting (5)]

MEMBERS ABSENT

   Dale Moberg (Axway Software)
   Youngkon Li (Individual)
   Kyoung_Rog Yi (National IT Industry Promotion Agency, Korea)
   Patrick Curran (Oracle) Leave of Absence
   Dmitry Kostovarov, (Oracle)
   Victor Rudometov, (Oracle)
   Hyunbo Cho (Pohang University of Science & Technology)

 [*: Voting member at start and conclusion of the meeting (0)]

APPROVAL OF PREVIOUS MINUTES
  The 2010-04-27 updated minutes were approved without objection
  <http://lists.oasis-open.org/archives/tag/201005/msg00008.html>

APPROVAL OF AGENDA
  The agenda was taken up without discussion
  <http://lists.oasis-open.org/archives/tag/201005/msg00007.html>

  Agenda:
  1- Review of comments (both external and internal) on the PR drafts
(closing May 9th).
    - TA Model specification (TAG-1, TAG-6), see proposals (email 5/5)
    - TA Markup specification  (TAG-1,2,3,4) see proposal TAG-4 (email 5/4)
  2- administrative matters: 
     - meeting notes approval (Apr 27).
     - JIRA issue tracker.
     - next PR.
     - promoting TAG work: venues, ideas, liaisons to other SDOs
     - future roadmap.


ACTION ITEMS

  AI1-May11: Jacques will check with Patrick Curran about when his leave
ends

  AI2-May11: Jacques will create a standing "TAG Promotion" item in the
agenda

  AI1-Apr27: [Dennis] to investigate "Compact" RelaxNG for our Markup
syntax.  [Jacques finds this tutorial:
http://www.relaxng.org/compact-tutorial-20030326.html]
  [Dennis since found a compact form used in a W3C specification, adding it
to TAG-4, http://lists.oasis-open.org/archives/tag/201005/msg00014.html]

  AI2-Apr27: [Jacques] to check the JIRA page, possible re-configuration,
add Dennis as authorized user.

  AI1-Mar30: [Jacques] follow-up with Dave P. on his comments, to make sure
we address them correctly.

  AI2-Mar30: *Done* [Jacques] Follow-up with Mary to add Dennis as new
secretary,
give access rights.

  AI3-Mar30: [Jacques] Contact W3C test team to remind them of PR, calling
for comments.

  AI4-Mar30: [Kevin] Contact Oracle internally to confirm they know about
the PR (folks who contacted him)

  AI5-Mar30: [Dennis] Investigate RDF possible representation for the TA
Model. (ontologies / OWL?)  




 1. REVIEW OF COMMENTS RECEIVED SO FAR ON THE PUBLIC REVIEW (ended
2010-05-09)
    We discussed the JIRA issues.

    TAG-1: Lack of Semantics in TAML.  We continued discussing the lack of
guidance on how to relate the markup constructs to the model and the
guidelines.  We are concerned about overloading the markup with semantics
and redundancy with the model.
  It looks like the semantics belongs more to the model than the markup.
  Rough proposal is to have the markup reference the model properly.  That
is the direction that we've started to take and we are still in rough
agreement.

    TAG-2: More-detailed concerns about TAML names and datatypes.  We
discussed that there are different levels to consider and how close this
looks to TAG-1 concerns.  This one is also about linking TAML to the model
better.

    TAG-3: Objection to Multiple inter-dependent documents.  We don't feel
that it is appropriate to make TAML a standalone.  We have to manage the
redundancy properly, in how we cross-connect.  It is still our commitment to
treat TAML as a representation of the model.

    TAG-4: Request for TAML Example and better
XML Skeleton Approach. 
   We agree this is valuable, especially having a better notation for
illustrating the XML structure without repeating the schema.  Our own
"friendly notation" is non-standard.  We looked at Relax NG compact
notation.  Jacques found a tutorial and there is a full spec.  The compact
notation seems good.  We can add the compact RNG (and also show it in the
example that we make, perhaps?).  [Note: Dennis found another simplified
scheme in a W3C spec since this discussion.]

    TAG-5: [Internal] Set up JIRA - Internal TC Task

    TAG-6: Broader Conformance Clause in the model.  Jacques reviews the
issue of needing statements of use and also demonstrating implementations.
Yet all are using the TA model and the guidelines and we need to give them a
way to be compliant users, such as the use by SCA.  Jacques has proposed new
conformance clauses.  Actual test-assertion instances should be cases of
use.  We will add new conformance clauses.  

    TAG-7: Bindings to test-assertion markup, and consider RDF and SPARQL.
Jacques says we could put this off and it is beyond what we ought to do now.

    TAG-8: Suggestion to specify RDFS and OWL for the model.  It seems not
critical to Jacques.

There's a new comment coming in that questions the use of plaintext for
predicates rather than having a machine-processable form.
  
 2. FURTHER PUBLIC REVIEW

We discussed the need for a second public review based on the changes that
we anticipate.  We will need to provide a change-tracked version for the
additional review.

On the May 25 meeting we want to make progress on proposed changes and see
if we are ready for another Public Review by the end of June.

 3. PROMOTION OF TEST ASSERTIONS

Jacques will create a promotion item in the standing agenda.  

Jacques talked about XML Daily and news about a government initiative, NIEM
and the naming and design and verification model.  They are setting up
standards to design XML schemas.  Jacques thinks those schema design rules
could be expressed as Test Assertions.  
  [Note from Dennis: There is a Federal CIO Council XML Community of
Practice (xmlCoP) distribution list that carries exchanges and contributions
on this topic.  CAM is being promoted for use with NIEM at
<http://sourceforge.net/projects/camprocessor/files/>.]

 4. DISCUSSION OF MAY 5 PROPOSED CHANGES

Proposes adding a semantic box, with color coding, etc. with the main
constructs.  Starts with a glossary, that needs to be matched to the formal
model and names, etc.  (Test Assertion is not in the glossary!)  [We have a
disconnect between the glossary and the formal terminology and need to work
that through.]  Jacques notices that we have nomenclature collisions with
markup and we need to deal with that.  Jacques thinks that normative
language should come out of the glossary and be in the formal part.  Also,
much of the descriptive material goes away and gets into the semantic boxes.

Talked about new conformance clauses.  Possibly merge classes 2-3, since
using sets could just be an option.  We still  need to work on how the
markup references the model, etc.

FUTURE MEETINGS

The next meeting is in two weeks - 2010-05-25.


Respectfully submitted,

 - Dennis E. Hamilton