Minutes, IBIS Quality Committee

21 October 2014

11:00-12:00 EST (08:00-09:00 PST)

ROLL CALL

Ericsson:                           Anders Ekholm
Intel:                              Michael Mirmak
IO Methodology                    * Lance Wang
Signal Integrity Software         * Mike LaBonte
Teraspeed Labs:                   * Bob Ross

Everyone in attendance marked by *

NOTE: "AR" = Action Required.

-----------------------MINUTES ---------------------------
Mike LaBonte conducted the meeting.

Call for opens and IBIS related patent disclosures:

Opens:

- None

ARs:

- Bob send partial source code to Lance.
  - Done.

- Bob locate source files defining IBIS_LINE_DELIMS.
  - Lance found this in ibis_chk.h.

- Mike recover percent format specifications in spreadsheet.
  - More progress.
  - Alignment problems for both columns and rows.
  - Manual checking required.
  - Bob: Atul is continuing to update the old spreadsheet.

- Mike add comments to CMPNT, DLY, MSPEC sections.
  - No new progress.

IBISCHK message list cleanup:

- Bob: I am working on CIRCUIT, Lance has a number to do.
- Lance: Some are self-explanatory, should we just copy the message.
- Mike: I'm not a fan of that.
- Bob: Sometimes nothing can be added.
- Mike: We could give an example message.
- Bob: For some we should just say what the check does.
- Mike: That is rewording the message.
- Bob: Examples would take more work.
- Mike: If we're not sure what might go into a %s we're not really explaining it.
- Bob: We need to do what we can with minimal work.

- Bob showed the message spreadsheet.
- Bob: There are questions about most RAMP checks.
- Mike: The question is whether to publish now or fix it up more.
- Bob: It is already published as a work in progress.
  - The next IBISCHK update may not have any new checks.
  - That may be weeks away.

- Lance: We had a "name too long - truncated" message.
  - We need to know what the limit is.
  - Sometimes it truncates names and that causes other problems.
  - What should we tell people?
- Mike: We should advise fixing the first problem first, then recheck.
- Bob: That is a good suggestion.

- Lance: What about "out of memory" messages?
- Mike: Users should know what those mean.
  - Bugs messages should simply be reported.

- Lance: There are a lot of empty comments in the EBDMDL section.
- Bob: You could say what the IBIS specification says about it.
- Mike: That would be for errors.
- Bob: For warnings we could say "This was not expected for CMOS", for example.
- Mike: Even looking up things in the specification could be significant work.
  - We should compare changes we make next time.

- Bob: Some bugs are actually errors, we should convert those.
  - For example Todd came up with an AMI file that gave a false error message.
- Mike: That is an error that is actually a bug.
- Bob: Where we find these we should report them.
- Mike: We should look for these as we go over sections, and make a list.
- Bob: Users should be reporting these to us.
  - We can just sort the list to go over the B checks.
- Mike: Some of these can be determined only by looking at the code.
- Bob: Some switch statements go to default because of some wrong input.

Travel plans:

- Bob: I am attending EPEP, will have to miss the next meeting.
- Lance: I am traveling Nov 5 to 27.
- We decided to next meet Nov 4

Quality related presentations for IBIS summit:

- Mike: The Ericsson validation presentation will be repeated.
- Bob: There are no others.

Meeting ended: 12:08

Next meeting November 4