Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[hyades-dev] Meeting minutes for 5/27 Hyades Execution Team Meeting

I've attached the meeting notes for today's meeting.

 

Kent Siefkes

 

 

Execution Team Group Weekly Meeting, 11AM Eastern, Tuesday, May 27, 2003

Attending
---------
Kent Siefkes, IBM Rational Raleigh
Ashish Mathur, IBM Rational Raleigh
Joe Toomey, IBM Rational Lexington
Kim Coleman, IBM Rational Cupertino
Richard Duggan, IBM Toronto
Dwayne Dreakford, IBM Rational Raleigh


Proposed Agenda
---------------
1. Review resource availability this week

2. Review Execution Use Case rescoping proposal for Rel. 1 (per meeting with Valentina and Marcelo on Friday), and approve (or modify)


3. Quickly review remaining task ownership discussed at the same meeting with Valentina and Marcelo (concentration was UI)


4. Discuss JUnit implementation approaches discussed yesterday (Dwayne, Richard, and Marcelo's email thread), and determine next steps



Recorded Discussion/Decisons
----------------------------
1. - Kent and Joe out-of-pocket in all day meetings, Wed-Thur, May 28 - 29
   - Dwayne out this Thur-Fri, May 29-30
   - Kris returns Wed, May 28

2. Agreed to the use case scoping for Release 1 as defined below (same as discussed with Valentina and Marcelo on Friday):

	IN (Priority 1): UC1 - Launch a single test instance from workbench on a single specified node
			[+ addition of creating a default ExecutionResult in test model with Inconclusive Verdict]
			[- prep for UC2 follow-on]

	IN (Priority 2): UC1.1 - Show control events to running test
			[Stop only]

   	OUT: UC2 - Monitor a running application on a specified node

	OUT: UC2.1 - Show control events to monitored app
 
   	IN (Priority 1): UC4 - Launch a test for immediate reporting/monitoring
			[UC1 plus Execution History]


3. On the execution side, confirmed Asish is signed up to do the Execution History loader, and Joe for the Test Tool (harness that takes the run test
   command from the UI and communicates with the various Execution Components to execute the test).

4. We agreed on the Marcelo/Dwayne approach to the remaining JUnit execution work:
   1. Write a generic JUnit runner, ie, a piece of code that knows how to run a JUnit test. This runner should be able to execute *any* JUnit test, sending a
      (Execution History) Test Event through the execution API for each basic JUnit event (test start, stop, failure and error).

   2. Customize the JUnit runner to recognize a Hyades JUnit test case, which will contain the additional information to map the JUnit tests to the objects
      in the Hyades model.  


   Next steps:
	a. Dwayne will add more comments to latest code (JUnit extensions) given to Richard, to aid transition back to Kris
	b. Kris will do the JUnit scheduler (runner), when he returns Wed.
	c. Richard will finish the IExecution Component factory for this weeks drop (make it more generic to do stub & skeleton mapping)
	d. Richard will work with Kris and Ashish to glue together the logging of Execution History events, through the framework, and into the loader Ashish
           is writing
	e. Ashish will work with Marius as necessary to write the ExecutionHistory loader
	f. Joe is going to do the "test tool", including the creation of the ExecutionResults object in the model


New Action Items
----------------
* Nothing beyond the next steps defined for #4 above.

Back to the top