From: Don Organ [don.organ@inovys.com]
Sent: Friday, May 31, 2002 11:07 AM
To: STIL. 4
Subject: STIL.4 teleconference meeting minutes 5/31/2002
Jim O'Reilly
Don Organ
Ernie Wahl
Dave Dowding
Tom Micek
Jim Mosley
 
 
 
Continuing from last week - regarding item #3 from http://65.119.15.228/stil.4/flowcontrolissues.htm
-------------------------------------
 
Dave - while I share Ernie's concerns that this must be "buildable", but I'm think we should be able to easily convey this to the user - therefore binning should be in the flow.
Ernie - if I have the binning associated with a test then it is easier to add and remove tests.
Dave - I see the bin as subserviant to a test
Ernie - each test has a pre-action, a pass-action and a fail-action (post-actions). You can do binning as the post-action.
Jim Mosley - I concur with Ernie - I don't see bins as just a terminal node. We need to have some sort of an expression to represent binning in a certain way - we need flexibility and expressions can provide that.
Jim O'Reilly - should we investigate binning strategies that have been used in various tester languages? I can drive this effort.
Dave - does anyone disagree that a Bin "is a part of the flow"? No dissention on this point.
 
Regarding #3 - binning will need further discussion. Other issues were agreed to (in general terms).
 
4. What are the return types from a Test Method (pass/fail, or pass/fail/error, or user-defined)?
Ernie - we must support parametric return values. Feels that "exception handling" should be phase 2.
Dave - I think error handling is something we need to support. I need to be able to distinguish between a pass/fail and an indication that the test was not successful (such as a search failure).
Ernie - why separate error handling from a failing condition?
Dave - there can be instrumentation errors (such as out of calibration) - and these errors can be "healed" (by recalibrating), but this requires control at the flow level. I feel pass/fail is very limiting.
We are all agreed that we need to have the TestMethod be able to return parametric results. We are all agreed that we need to be able to return pass/fail. We are not all agreed that the error return needs to be explicit. (we avoided dealing with what "user-defined" means).
 
 
 
Tom - regarding multisite: Conference calls on Tuesdays at 3:30 Central time (1:30 pacific time). (888) 499-3772. Passcode 709880#.
 
 
 
-----------------------------------------------------------------
Don Organ                      Inovys Corporation
(925) 924-9110 x122
-----------------------------------------------------------------