Minutes meeting Feb 14

["Durand, Jacques R." <JDurand@us.fujitsu.com>]

For review:

 

 

Event Description:

------------------

TaMIE TC conference call

http://www.oasis-open.org/apps/org/workgroup/tamie/

 

Date:

Thursday 3pm PT(California time), 14 February 2008

Host: Fujitsu

US toll free: 866-760-1934

intl: 1-517-233-9562

participant: 2426197

 

Agenda:

-------

0. Minutes of previous initial meeting.

http://www.oasis-open.org/apps/org/workgroup/tamie/email/archives/200801/msg00002.html

1. "Transfer" of eTSL input doc (already donated to IIC TC) to TaMIE. Do all authors agree.

2. Presentation of the main input document (eTSL)

- does it fulfills the objectives of the charter.

- about separating the event model & management, from the test scripting and exec model.

3. Discussion on the deliverables: Requirements, Evt Model, Test Scripting. COntent and Timing

4. F2F and other practical discussion.

 

Attendees:

----------

Members present:

- Mohammad Abidi (AIAG)

- Hyunbo Cho (Pohang Univ., Korea)

- Jacques Durand (Fujitsu)

- Sangwon Lim (KIEC)

- Nenad Ivezic (NIST)

- Stephen Green (SystML)

- Chuck Morris [observer]

Excused:

- Dale Moberg (Axway)

- Youngkon Lee (KIEC)

 

Action Items:

-------------

AI1: Jacques to send out a "template of donation" letter, for eTSL contributors

to send to TC and its chairs.

AI2: Dr. Cho to provide contact of Jungyub Woo (Samsung) to make sure he donates

his contributions.

AI3: Jacques to upload the eTSL draft on TaMIE page.

 

Minutes:

----------

0. Minutes of previous initial meeting.

http://www.oasis-open.org/apps/org/workgroup/tamie/email/archives/200801/msg00002.html

- Minutes approved by TC.

 

1. "Transfer" of eTSL input doc (already donated to IIC TC) to TaMIE. Do all authors agree.

 

- Jacques: the eTSL doc must be "officially" donated to the TC.

It has already been doanted to former OASIS IIC TC,

but authors need to confirm they donate it again to this new TC, according to its IPR mode

(which is the same as IIC).

- All authors or org reps present, confirm orally they do not have any problem donating this IP.

- Jacques to send out a "template of donation" letter, for eTSL contributors

to send to TC and its chairs.

 

2. Presentation of the main input document (eTSL)

- does it fulfills the objectives of the charter.

- about separating the event model & management, from the test scripting and exec model.

 

- Jacques presented eTSL from a high level.

Motivation was:

- portability of test cases across test environments,

- provide a precise scripting for test cases - still higher-level than programming lge - with

unmabiguous semantics.

- usable beyond QA testing: Should be usable as a "concrete, executable way" to specify

message-based B2B contracts between partners, or to implement higher-level contracts

(e.g. choreography dialects, collaboration definitions)

- platform-independent XML markup, that leverage existing XML tools and dialects (XPath...)

The scope of test cases has the following emphasis:

- ability to test the integration of several B2B standards, and beyond WS-I objectives,

integrating also user-level

standards (domain-specific, business documents...) and profiles.

The operational requirements were:

- versatility: both testing (conformance, interop) and monitoring, e.g. for SLA follow-up.

- ability to test/monitor either in real-time (e.g. to generate alarms) or in deferred mode

(e.g. over a log of B2B transactions)

The challenges we will face:

- splitting the event management model deliverable, from the test scripting and execution model.

- a scripting mark-up that leverages external "specialty dialects" that are widespread enough

(e.g. XML-based) while providing a minimal set of native primitives to express the workflow

of test cases, that can easily and unambiguously be implemented. Idea is to not reinvent

existing dialects, but current wkflow languages are too sophisticated on the flow control

side, and not enough on the event management side.

- an event model that provides XML wrappers for any type of data to be used by test cases.

 

3. Discussion on the deliverables: Requirements, Evt Model, Test Scripting. COntent and Timing

 

- The "requirements" deliverable: need not be completed before starting work on spec deliverables.

This document will serve also as a way to communicate what kind of problems TaMIE is addressing

(so used to illustrate the puspose of TaMIE work as much as a requirement foundation for us.)

Indeed, requirements at some level will need be adjusted to what we actually can resolve.

- Nenad: Requirement doc should contain a collection of concrete B2B Use Cases.

- Jacques: yes we have a bit of this at the end of eTSL draft (section 6). Need more.

Beside Use Cases, requirements can also be more directly assertive - see

Intro parts of eTSL (sections 1,2)

- Event model and Test script model deliverables probably need be designed in parallel,

yet will try to keep separate.

Tentative date end 1st half 2009 for a set of specs ready for public review.

- The implementation deliverable: will not pretend to be a "refence implementation",

but mostly a proof of concept, as in fact it is required by OASIS that some implementation be

done in order to pretend to OASIS standard status.

Will also help detect flaws in the specs before they are complete.

KIEC members volunteered to lead this implementation work that will take place in parallel

(they already have an implementation of eTSL draft,

and will provide feedback, according to Cho).

 

4. F2F and other practical discussion.

 

- next meeting: Thursday Feb 28th, 3:00pm-4:30pm PT (starts 1h before this meeting)

emphasis will be on use cases to be addressed.

- F2F: so far, plans for March 31 - Apr 1 in Bay Area (airport: San Jose),

Fujitsu offers to host.