Michael Mirmak convened the meeting 17 NOV 2021.
Randy Wolff took minutes.

Attendees:
ANSYS                   Curtis Clark
Intel Corp.             Michael Mirmak*
MathWorks               Walter Katz*
Micron Technology       Randy Wolff*
Siemens EDA             Arpad Muranyi*, Weston Beal
Teraspeed Labs          Bob Ross*

No patents were declared.

After a review of the minutes of the 03 NOV meeting,
Arpad moved to approve the minutes.
Randy seconded.
The minutes were approved.

actions status:
  [AR] Bob will report a bug to the parser developer about the 2 space
  characters.
  This is done.

  [AR] Randy will add to the known issues list:
    - relax requirement that all pins be listed in [Designator Pin List].
	- define "alphanumeric" as used in the IBIS specification.

  This is done.  Randy noted the issues are in an "Editorial" list, but he noted they require BIRDs.  These items might be better suited to the Interconnect task list when that group meets again.
 
Arpad commented that we should advertise the IBIS 7.1 draft better on the website. 
[AR] Randy to request Steve to post a link to the IBIS 7.1 Draft directory in a banner on ibis.org


agenda:
Walter suggested about alphanumeric.  SPICE has constraints on node names, model names, subcircuit names, etc.  There is a subset of the alphabet, numbers, and some symbols that are legal for all of those fields.  We could talk about an SI character set.  Randy noted we have these descriptions in IBIS-ISS.

C Comp Model Discussion:
Bob noted the [C Comp Model] keyword descriptions had incorrect sentence alignment.  
[AR] Michael to fix alignment in a Draft 20.

Bob described the issue with [C Comp Model].  He noticed a check was missing in the latest IBISCHK7.1.0 code for [C Comp Model].  For example, the following terminal lines in a C Comp Model should cause an error from the parser:

4    Pullup_ref   
5    Pullup_ref   

Randy said it was his intention in BIRD200 that terminal numbers should have distinct Terminal_types, so that there are no repeated connections to the available Terminal_types (Buffer_I/O, Buffer_I, Pullup_ref, Pulldown_ref, Power_clamp_ref, Gnd_clamp_ref, Ext_ref, and A_gnd). The syntax was not intended to allow an IBIS-ISS subcircuit to have lots of nodes, and then to use the terminal lines in C Comp Model to connect the various rails multiple times.

Randy also noted the A_gnd had this statement that is allowing A_gnd to be defined multiple times:

If present under File_IBIS-ISS, Terminal_type A_gnd may be used any number of times on any of the terminal lines.

There was discussion on whether A_gnd should be allowed more than once. There was agreement that it should occur only once too.  

The following statement was added:

"Any given Terminal_type may appear only once in any C Comp Model."

The following statement was removed:

"If present under File_IBIS-ISS, Terminal_type A_gnd may be used any number of times on any of the terminal lines."

Arpad noted that this statement was poorly written.  It can be read as saying that "A_gnd" could be repeated multiple times on a terminal line.  Randy said this would fail a syntax parser but agreed it was poorly written.  Randy found that this text was copied from the Interconnect Model section.  Arpad felt this should be fixed in that section.  Randy thought this could be added to the Known Issues document started for IBIS 7.1.

[AR] Randy to add the A_gnd sentence issue to the IBIS 7.1 Known Issues document and send it to Bob for posting.

[AR] Michael to finish editing of Draft 20 and send it out for posting.

Randy Wolff moved to adjourn.
Arpad Muranyi seconded.
The meeting was adjourned.