The next meeting of the VASG will be in Orlando, Florida in conjunction with
the Fall VIUF. I will send a separate email that covers the exact date and
time as well as agenda. The meeting will be either Wed or Thurs (6th or 7th
of Oct); I am working to confirm with the conference organizers.
Second:
Friday, 17 Sep 99 is my last day in the service of VeriBest. I start work
for Synopsys on Monday, 20 Sep. I do not yet have a Synopsys email address.
I can always be reached at the following:
stephen@srbailey.com
303-588-2001 (mobile phone number)
Third: The Minutes
Minutes of the VASG Meeting of 3 Sep 99, Lyon, France
Attendees:
Alex Zamfirescu
Victor Berman
Jean Mermet
Alain Vaichoux
John Hillawi
H H Kamal Hashmi
Jim Heaton
Dennis Brophy
Peter Ashenden
Steve Bailey
Steve presented a hand-full of slides reproduced below:
Agenda:
* 1076a, Shared Variables, Status
* 1076-99 Status
* PLI Status
* Discussion on scope/guidelines for VHDL 200x
* Next Meeting
* AOB (there was none)
* NOTE: The minutes present the information in the same order actually
presented at the meeting which differs from the Agenda order.
1076a Status [summarized]
* Our RevCom mentor has worked with various folks in the IEEE and has
identified an acceptable process for completion.
* The RevCom mentor will present this proposal to RevCom at their next
meeting (15 September, I believe).
* See attached email from RevCom mentor (James Moore) for details.
* I (Steve Bailey) will let the WG know the RevCom response to the
proposal.
1076-1999 Status:
* Due to 1076a issues, no significant progress has been made on
1076-1999
* Paul Menchini will provide a draft 1076-1999 document for review in
time for the Fall VIUF
* Assuming that RevCom approves the 1076a proposal for closure
* Otherwise, Paul may have additional 1076a work to perform
* The chair has been late in starting the process of forming a ballot
constituency
* With the 1076a issues, probably best to wait until after the Sep.
RevCom meeting prior to proceeding
PLI Status:
* No update was received from Francoise.
* PLI committee continues its regular meetings and they are
progressing
Next Meetings:
* Fall VIUF in Orlando, FL.
* Probably Thursday, but may try to fit meeting during evening on Tues
or Wed. Anyone with an opinion should share it with the chair.
* Whenever the next DASC steering committee meeting will be (probably
HDLCon 2000).
VHDL 200x Guidelines
* The chair asked for volunteers to draft a guidelines document.
* Victor Berman volunteered
* The chair also believes that Paul Menchini would likely volunteer
and took the action item to discuss it with Paul.
* The chair set out the following non-controversial (at least at this
meeting) proposed guidelines:
* VHDL 200x will avoid duplicating the efforts of other groups without
the consent of and coordination with those groups. The groups specified
(but not limited to) include:
* OOVHDL
* SID
* 1076.1
* Others as deemed appropriate
* It was also noted that this does not preclude the consideration of
incorporating 1076-related standards directly into 1076-200x (specifically
1164).
* Additionally, the chair discussed the following proposed guidelines:
* Maintain backward compatibility whenever possible except:
* New reserved words can invalidate models using those words as
identifiers.
* Any "bug fix" to the language that is necessary and would break
existing models that rely on the "erroneous" language specification would
not be compatible, but would fall within the scope of 200x.
* Any significant changes to existing language capabilities that could
result in existing models not operating correctly can be incompatible as a
way to help users flag and change this functionality. (An example of this
situation would be the VHDL '93 TextIO changes that were not backward
compatible by intent.)
* Keep the problem domain (language scope) targeted at electronic
systems design and description.
* All new language features and capabilities must be consistent with
the existing language design (conceptual integrity of the language).
* EDA Vendor support shall be promised for all proposed changes.
* The attendees pointed out to the chair that this guideline will be
difficult, if not impossible, to implement and that it is a different class
than the other guidelines.
* The chair stated that the intent is to ensure that we significantly
increase the likelihood that only language changes that are likely to be
supported will "make the cut" for VHDL 200x. In that spirit, the chair is
open to suggestions as to how to help ensure this happens.
* Victor Berman will research the information from the North Carolina
workshops for incorporation into the draft guidelines.
* The chair will find a copy of the VHDL '92 guidelines document.
Alain Vaichoux suggested sources for this information.
The only additional discussion during the meeting that the chair would like
to specifically identify is a concern expressed by Victor Berman about the
survey effort currently being performed by the folks under contract to the
USAF for the VHDL 200x work. Victor thought that it could be harmful if we
end up with a laundry list (my words, not his) of requirements that the VASG
feels it must treat each request equally. His concern was expressed in the
context of the discussion on EDA vendor commitment being a guideline for
VHDL 200x.
In this regard, this may be the most difficult aspect of the VHDL 200x
process: Ensuring that we address the high priority, high return issues as
these are the ones that are most likely to win quick EDA vendor support.
How and whether this can be effectively codified as a guideline is a
challenge for Paul and Victor.
Stephen Bailey
Chair, VASG (1076) and 1076a.