From: Francoise Martinolle (fm@cadence.com)
Date: Mon Nov 03 2003 - 12:33:32 PST
Minutes of BTF Meeting November 3rd, 2003
Meeting started at 10:40 Pacific Time
In attendance:
Don Mills
Cliff Cummings
Kurt Baty
Mehdi Mohtashemi
Francoise Martinolle
Tom Fitzpatrick
Karen Pieper
Ennis Hawk
Stuart Sutherland
Brad Pierce
Steven Sharp
Stefen Boyd
Kathy McKinley
Francoise Martinolle agrees to take minutes for this meeting, with the
understanding that she is not volunteering to do so for all meetings.
We need to establish some kind of rotating responsibility for this.
Fitz moves to accept minutes from October 20. Mehdi seconds. No opposed,
no abstain, passes.
* Guideline group report
Jay does not have the time to continue driving this group, replacement
chair is needed.
The charter of the guidelines group was defined and was mailed to the BTF
to be reviewed and accepted.
Kathy volunteered to chair the guidelines group. Approved.
Kurt described the three deliverables expected from this group:
1. a process flow specification:
This will be written document by Kurt which will provide codification
of the methods and the procedures to follow to create language enhancements.
2. extension guidelines specifications.
These specifications will describe the spirit of Verilog (how Verilog
works, key terminology used in talking about the language). Stuart
Sutherland's help will be appreciated.
This document will describe how language constructs are meant to be used
and will provide a high level architectural design for the extensions. This
document will be used to coordinate the several groups which will be in
charge of adding new language features so that they are consistent with the
spirit of the language and coherent with the language itself.
Serves as a reference for an approach to extend the Verilog language.
Backward compatibility, concise clarity in the language are examples of
such guidelines.
3. a proposal evaluation template
1 or two pages long, it is a check list sheet which needs to be check out
and filled out by someone making a proposal.
It will help preventing issues to be dealt later.
The group will collect questions to be added to the template.
Steve Sharp said that he only received 1 comment from Mac on the charter of
the guidelines and Steve asked for more comments on the charter of the group.
Tom Fitpatrick: a guideline is not a rule.
Ennis had time concern: how long is that going to take?
Kurt estimated 2 months at least to come up with a draft.
1st document can be expected in the next 4 weeks
2nd document: an intermediate document can be shown and
circulated to the BTF at the next meeting
1st week of January , we should have a draft of these
guidelines
Kathy proposed to have a rough schedule in 2 weeks to
be provided to the BTF with milestones
Important factors: useful for people bringing proposals , and also to
provide high level view of what we trying to do here.
Volunteer on the templates: Anu
Stuart commented on the charter: not confident that it is going to be used.
Brad would like to see something about actual implementation in the
template sheet.
Motion to approve the charter: Kurt moved, Second: Stephen
abstain: Ennis, Karen
no opposed. Passes
Members of the EGG group are currently : Steve, Kurt, Alec, Tom, Anu, Kathy
Any specific guidelines can be mailed to the reflector.
* Alec's Prioritization group:
No one thought that they met again.
Steve to ask Alec for an update and ask this group to start prioritization
of the donations.
Steve will ask Alec if he is done or if he is going to have another meeting.
Kurt is expecting more details on how the donations depend on each other.
Accellera Donation is coming (data types fairly stable), Jeda donation is
also overlapping,
Cadence has donated a data type specification.
Kurt wants orthogonality in the language.
Alec had proposed the formation of several subgroups of the BTF.
Cliff: datatypes should be started
Kurt: we should do packages first, a lot of other things will become easier.
Kurt would like to see packages and overloading.
There are a few subgroups which have the least conflict with systemVerilog
Steve Sharp asked what was the legal situation with starting to work on
systemVerilog? Concepts are in the public domain. we can discuss the
concepts, but we should wait for an agreement between IEEE and Accellera to
begin consider the detailed SV proposal.
Ieee and Accellera should work out what can be done.
Directions should be given from the VSG.
Steve Sharp will add this topic to the agenda of the next VSG
meeting.should be asked directions
IP protection sub group:
Mac was going to ask for IP vendors to participate (ARM)
task forces for subgroups,
The assumption is that task forces for sub groups will develop a proposal
first, produce the LRM changes or enhancement, get BTF feedback,
perhaps further improve the language, the the BTF will vote on the changes
and move it to the VSG where it may get further reviewed, clarified and
finally passed.
The process codification should clarify how this is done.
Sub task groups will be formed for behavioral enhancnement under the BTF.
There may be a different group working on assertions.
Motions will be made in the BTF to create subgroups.
* Surveys
Are we done with the survey? We may be capturing more data in the spring
(Dvcon) if necessary.
Kurt thinks he is done with the survey work.
Motion to ajourn: Cliff moved.
Next meeting the 17th of November.
This archive was generated by hypermail 2.1.4
: Mon Nov 03 2003 - 12:26:42 PST
and
sponsored by Boyd Technology, Inc.