Attendees: *Gerald Chan, *Ric Dokken, *Scott Franzen, *Carey Garrenton, *Greg Maston, *Jim O'Reilly, *Ernie Wahl

Line and clause numbers if any are with regard to STIL.4 syntax document D02 dated October 24, 2016.

================================================================================
OLD BUSINESS:

1. Clause "32. TestType Definition":
    REPLACE: For each standard-defined test-type, STIL.4 or STIL.5 shall describe its function and how it uses its parameters to perform that function.
    WITH: Each standard-defined test-type shall be accompanied by a description of its function and how it uses its parameters to perform that function.
          That description should enable replicating test results across various ATE with identical electrical properties.

2. Insert in clause "21.1 Literal Values" after 2nd paragraph (rewritten from May 9, 2016 doc, Line 662):
    User-data AMPERES, CELSIUS, DECIBELS, DEGREES, FARADS, HENRIES, HERTZ, METERS, OHMS, SECONDS, VOLTS, and WATTS represent literal values with appropriate units, e.g., 100mA for AMPERES. PERCENT represents an integer or REAL followed by a percent sign, e.g., 5%. REAL represents a floating point number without units, e.g., 0.05.

3. Add "Node" to clause "3. Definitions":
    used to describe a node in a directed flow graph (flow-node) or an electrical node in a circuit

4. Syntax BNF inconsistently uses ::=, :=, and =, 7/5/2016 consensus: use ::= for BNF

5. Meta-type "next_stmt" is not defined in Syntax Document

6. Review Annex D: determine whether it's appropriate for this standard. Goal is to have ONLY informative Annexes.
================================================================================
DISCUSSION:
    
1. With regard to form 3 Test instantiation syntax: since whether a TestMethod
is an object, function, or something else is unspecified, hence MethodParameters
can not be reliably interrogated from outside the test instantiation:

  a. Group consensus is that the syntax and description is sufficient to convey
     that fact.

  b. That is why MethodParameters do not support directionality "Out".

2. TestType and TestMethod names occupy separate namespaces.

3. Except FlowExtended standard definitions, black-box test-types and
test-methods are not defined by STIL.4.

4. With the acceptance of form 3 Test instantiation syntax, Ernie would like to
add FlowExtended TestBase syntax statement:

    TestExec TestMethod method_name;

TestMethods can then be used for building user-defined tests.
================================================================================
CONSENSUS:

1. New 3rd form of Test instantiation syntax, example, and semantic description:
    1 Test DSPtest {
    2   TestMethod DSPmethod;
    3   MethodParameters {
    4     In sigref_expr sigref = 'AnalogOut';
    5     In Frequency sampleRate = 200kHz;
    6     In Decibels SNR = 53dB;
    7     In Percent THD = 0.2%;
    8   }
    9 }
    Proposed 3rd Form of Test Block:
       Test TEST_NAME {
        TestMethod METHOD_NAME;
    	( MethodParameters {
    	( <In | InOut> param_type PARAM_NAME = expr;)*
    	})
        }
    
    
    The third form creates a Test block with parameter assignments and a
    reference to a TestMethod name.  This form allows for connection between a
    Test block and a named TestMethod with optional passing of parameter data.
    TestMethods are not defined within STIL.4, and therefore define before use
    does not apply to the test method name.


2. Insert new figure for Test after figure 60 as per Ric's email Fri 11/4/2016
   4:16 PM, test part of attached file Flow_FlowNode_Test.pdf

   Add figure caption: Figure 61 —Diagram: Conceptual Model for Test in Flow basic
================================================================================
ACTION ITEMS:

Ric: CONSENSUS item 1 - 2

================================================================================