From: Don Organ [don.organ@inovys.com]
Sent: Friday, March 08, 2002 12:55 PM
To: don.organ@inovys.com; dave_dowding@agilent.com; bill_chown@ims.com;
daniel@san-jose.tt.slb.com; wahl@aloft.agere.com;
Gordon_Robinson@3mts.com; gwilder@dal.asp.ti.com; gmaston@qwest.net;
h.ehrenberg@att.net; jason_doege@inovys.com; mosley@san-jose.tt.slb.com;
jim_oreilly@agilent.com; jim.showman@inovys.com;
jose.santiago@philips.com; larry.moran@teradyne.com;
stuart.h.nelson@teradyne.com; Tom.Micek@motorola.com; tonyt@synopsys.com
Subject: Corrected: Minutes from STIL.4 meeting 3/8/2002

Here's a corrected copy of this morning's conference call. Please delete the earlier copy.

Thanks.

-DVO-
---------------------------------------------------
3/8/2002 - conference call from 10-11am Pacific Time.

Attendees:
Dave Dowding, Tom Micek, Ernie Wahl, Gordon Robinson,
Jim O'Reilly, Daniel Fan, Tony Taylor

Don Organ volunteered to keep minutes.

Dave is trying to get our own e-mail reflector.

Agenda:
    Go through attachments (requirements generated since last meeting)
    High-level walkthrough of the Inovys proposal.

Ernie's requirements (were distributed via email before the meeting)
    Object-oriented requirements gathering
    Gordon: STIL is declarative - and this is important
    Finite State Machine behavior (be able to jump from any place to any other place):
        Ernie - within a scope (block)  should have FSM capability - based on test results. Gives greater flexibility.
        Gordon - is your description really a FSM, or a general purpose procedural language?
        Ernie - state includes the complete history of what has been done
        Daniel - is it possible to control the bin as you go through the flow?
        Ernie - certainly
        Gordon - at what point do we stop adding computational embellishment and referr to an existing/known programming language?
        Daniel - can not assume just pass or fail in determining flow
        Ernie - discourages N-ports exits (after porting to many testers) - recommends pass-fail for most porting. Not all machines can do FSM.
        Dave - we should add FSM to a glossary (Ernie to provide a definition)
        Dave - lets move on - (we are just reviewing these ideas now - not accepting/rejecting requriements)
    Maintenance Conscious design desicions - for example - should be easy to add/subtract/change speed binning (perhaps changing the program by writting a script)
    Pass of prior test results to subsequent teset parametners - for example, a limit may be dependent on prior test result
    Conditional test/testblock execution (based on previous results - or based on package test versus wafer probing)
    2-phase development - for generic spec, and for targetting a specific tester

Jim Mosley's requirements
    #1) ability to reuse tests and subflows
    Gordon - it is hard to understand what exactly is implied by the Goto statements and the subsequent binning algorithm.
    Daniel - flow is sepearated from the binning
    Ernie - haven't yet drawn the distincition between a node and a test
    Dave - let's contrast this with Don's example

Don's overview
    Ernie's question about binning
    Don - this doesn't address the binning issues raised in Jim Mosley's requirements


Dave - we'd like to get an example we all feel good about.
Gordon - let's get more examples - that we help raise the right questions
Ernie - Don's OddEven.stil may be an example of the FSM approach.