Meeting today... 2pm PT

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

Title: Meeting today... 2pm PT

All:
Sorry for late notice:
(and also I was told I sent wrong minutes last time: Feb 24 last year! Here corrected)
Next call:
Time: Monday March 8, 2004, 2pm PT
Host: Fujitsu
Toll - : 1-512-225-3050
Passcode: 89772#
Agenda:
1. Test Case scripting: Status on remaining issues, summary of updates (Mike Kass, Jacques)

- review of recent issues.
- requirements for bus process testing (Tim): correlation issue.
- Current state of spec doc (content, status).

2. Test Framework interfaces:
- Status on standard "astract" interface/API Test Service - MSH.
- Can we settle on an envelope-agnostic "message wrapper" document (header fields + payloads)
3. Other:
- Test harness architectures: the Korbit case for remote conf testing.
- Update on ETSI European test event, need for pre-testing.
- IIC f-2-f New Orleans April 28-29 details: focus on the agenda.
Regards,
Jacques
<<IIC_February_23_04_minutes.txt>>

IIC Conference Call Minutes: 
Monday, February 23, 2004
 
Attendance:

Mike Kass (NIST)
Jacques Durand (Fujitsu)
Serm Kulvatunyou (NIST)
Monica Martin (Sun)
Steve Yung (Sun)
Tim Sakach (DrakeC)
Pete Wenzel (SeeBeyond)

Excused:



Minutes taker: Jacques Durand

Time: Monday February 23, 2004, 2pm PT
Host: Fujitsu
Toll - :  1-512-225-3050

Agenda: 

1. Test Case scripting: Status on remaining issues, summary of updates (Mike Kass, Jacques) 
- Try to wrap-up on the control statements (split / join, if-then-else). 
- Clarify the "exclusive thread" issue: XORjoin vs using some signal variable.
- Try to wrap-up on semantics of GetMessage of the message store, especially in case 
of multiple mesges (see use case). 

2. Test Framework interfaces: 
- Status on standard "bastract" interface/API Test Service - MSH.
- Can we settle on an envelope-agnostic "message wrapper" document (header fields + payloads) 

3. Other: 
- Update on OAG meeting, ETSI European test event. 
- "light" IIC f-2-f New Orleans April 28-29 details: focus on the agenda. 


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

1. Test Case scripting: Status on remaining issues, summary of updates :

1.1 Control statements:
- split: only need one "split", forking threads asynchronously (if just one, 
asynchronously with main thread). Any synchronous control can be obtained by "join",
e.g. if we want to fork "synchronously" to the main thread, we'll join at the next step.
- "And-join" and "or-join" are both needed. 
- if/then now started from within any test step (at Assertion level) can start threads.
It seems we do not need at higher level.

1.2 exclusive thread" issue: XORjoin vs using some signal variable:

- "Xor-join" does not have a clear use case so far. 
Would force to wait for all joined threads to timeout, which may be impractical.
Also would not help the use case where one wants to detect not just exclusive threads 
at completion, but simply exclusive events (e.g. if one gets a message "Accept" one should
not get a message "Reject"). For the latter case, one can start threads that act as
concurrent Exception threads (catching messages that we are NOT supposed to receive).
- in any case, we still need to define more precisely the notion of "thread completion":
if the last step terminates by "failing" the test case, this is anyway an exit condition.
if the last step terminates on an "undetermined", is it also an exit condition?

1.3. Semantics of GetMessage()

- GetMessage filter: operate on the aggregate of (unmasked) messges in a store.
- Will execute as many times as the store is updated, untile stepDuration expires or 
until the filter succeeds.

1.4. Other:
- (Tim) Correlation of incomingmessages: when the correlation attribute value is 
known statically, can be hardcoded in test case. But we may not know the correlation value 
in advance. It seems we need "parameterized" XPath expressions. Tim will investigate.

2. Test Framework interfaces: 
- Mike sees this wrapper as same obejct we pass to the "Initiator" action.
That should work.
- However, this wrapper should be indepdt from ebMS format: apps as well as Test Service
are not expected to parse - or to generate - any ebMS header. In addition that would
induce more ebMS dependency (V#, etc.)


3. Other: 
- Tim says Drake C. is getting its test harness ready for process level.
Steps are: handle combination of BODs and ebMS, use D.C. TestFramework,
then write test scripts, then apply to interop testing (use Accordaire Reflector to 
monitor and validate message content.)
- OAG-NIST testbed to be used in IV&I pilot. This would involve multiple types
of testing, and business process level. Cjhallenge: Vendor , and STAR adoption.
- "light" IIC f-2-f New Orleans April 28-29 details: we have Thu 29 from 1pm to 5pm.
We'll have a 1 or 2 hour  meeting. The theme would be more "higher level" than
technical:
. current feedback on test framework implementations, their usage (e.g. for ETSI round).
. perspectives in using T.Fk for other OASIS specs.
. the role of IIC if a new ebMS arch or coordination group emerges.
. role of IIC in emerging testing initiatives based on IIC test material.
. prioritization of work directions, due to limited bandwidth.