From - Mon Jul 27 13:53:35 1998 Return-Path: Received: from commx ([153.33.32.81]) by mail1.ltx.com (Netscape Messaging Server 3.01) with SMTP id AAA10622 for ; Wed, 8 Jul 1998 15:48:32 -0400 Received: from tessi.com by commx.ltx-tr.com; Wed, 8 Jul 1998 12:53:43 -0700 Received: from ihgw2.lucent.com by tessi.com (4.1/SMI-4.1) id AA08558; Wed, 8 Jul 98 12:43:01 PDT Received: from aloft by ihig2.firewall.lucent.com (SMI-8.6/EMS-L sol2) id OAA05388; Wed, 8 Jul 1998 14:34:53 -0500 Received: from kiwi (kiwi.micro.lucent.com) by aloft (4.1/DCS-aloft-M11.3) id AA27270; Wed, 8 Jul 98 15:45:58 EDT Received: by kiwi (SMI-8.6/DCS-aloft_client_sv-S2.5) id PAA14795; Wed, 8 Jul 1998 15:44:11 -0400 Date: Wed, 8 Jul 1998 15:44:11 -0400 Message-Id: <199807081944.PAA14795@kiwi> From: Ernie_Wahl To: stil-a@tessi.com Subject: STILA4 - Flow Control Meeting Summary X-Mozilla-Status: 8001 Participants: Spass Stoiantschewski Tony Taylor Jose Santiago Jim Felte Ernie Wahl We completed the walkthough of the Envision/TPG3 comparison material at http://www.ltx-tr.com/STIL/A/group4/index.html The comparison included: 1. testflow object inheritance 2. testflow containment structure 3. test structure (general and ac test example) 4. good device binning We can now draw from two models to design the datamodel for the STILA testflow. There appeared to be agreement that we need the datamodel before any meaningful progress can be made on syntax. There are questions standing in the way of progress, begging to be answered. We agreed that they should be posted so that those not with us today would have a chance to ponder and/or offer alternatives: 1. Should the STILA testflow: a. represent an entire test program ? b. be a database from which to generate a test program ? 2. Should we use the inheritance hierarchy from: a. Envision ? b. TPG3 ? c. some combination ? 3. Should we use the pre/body/post test model ? 4. Should binning objects be: a. independent ? b. associated with a test/testblock ? c. both ? 5. Is the good device binning representation acceptable ? Are there alternatives ? 6. Are multi-binning tests acceptable ? The intent is to establish a system of constraints that will allow us to define a testflow and tests without having to consider every permutation of every good idea. Whether and which of these constraints are acceptable should become evident once we consider some testing scenarios. Well, that's the theory ... ============================================================================== Ernie Wahl Department 538240000 Computer-Aided Test Lucent Technologies Room 22y-210en 555 Union Boulevard email: ejwahl@lucent.com Allentown, Pa. tel: (610) 712-6720 18103-1229 fax: (610) 712-4235 ==============================================================================