Meeting 41 Minutes

Teleconference, March 10^th , 2008

A quorum of nine out of eleven voting members present. One non-voting member is also present. Attendance and voter status indicated on the roster web page.

Attendees

• Voting members: Charu Aggarwal, Serrie Chapman, Sandeep Desai, Yaron Kashai, Andrew Piziali, Darren Galpin, Joseph Hupcey, Michael McNamara, Mark Strickland

• Non-voting members: Joe Daniels

1. Call to Order

Call to order on March 10th, 2008 at 9:04 am PDT (1704 UTC).

Andrew reminded members to mail a message to “rollcall@ieee1647.org” in order to record their meeting attendance.

2. Approval of Meeting 40 Minutes

Andrew pointed out the time of today's meeting was reported as 1700 UTC rather than 1600 UTC because of the change to daylight saving time in the U.S. He said he will correct this while posting today's minutes.

Serrie Chapman made a motion to approve the meeting 40 minutes.

Yaron seconded the motion.

Motion approved.

3. Status Update

Andrew : I’m pleased to report that the P1647-2008 ballot concluded on March 6th with a success—in no small part due to the work of the working group in putting together a document that passed the eyes of 24 of the 29 members of the ballot group. There were 24 affirmative votes cast. There were no negative votes so that’s an 82% percent return rate that meets the IEEE Standards Association 75% of the ballot group casting a vote. There were seven comments included, among those all but one were small editorial corrections. The one comment that was not an editorial correction, I was rather embarrassed to read because it was a person in the original '06 ballot group who asked for a small change in a method which calls an external program from an e program allowing a message and the return code to be captured and used. According to that person it was promised to be corrected in the next revision of the draft but I was not aware of that. The reason that it slipped through was that an issue had been opened but that issue had been closed and so was not in a deferred state and we did not review all of the closed issues in the early 2008 draft standard process. Needless to say, I am opening an issue for that and that will be considered for the next time around, it is a nice thing to have but we did kind of promise this person that it would be in there.

Yaron: - do you recall the name of the person?

Andrew: Yes, I have the ballot results up here—that was um …Piotr Karocki. When I look at the comment in particular this has to do with a method called fatal(). Fatal() apparently currently takes no arguments and he suggested that it take status as a message string and that way you could satisfy this information and in this way it would be the termination of any program. That one issue aside, everything has gone very smoothly through the ballot process and thanks to all who are in the ballot group and voted. Because there are only editorial corrections and one suggestion and no negative votes I don't expect there to be a need for a recirculation ballot. I will be touching base with IEEE later on this week to confirm that, and to coordinate any final editorial changes that do need to be made prior to publication. Any discussion on ballot or results?

Joe H: Are we signed, sealed and delivered after Joe D. makes a couple of these changes? What happens then?

Andrew: Yes. With respect to moving forward we’ll be reviewing the comments, open issues for Joe to work from to make editorial changes that need to be made. We hope to finish that this month. There is still some code sample formatting that Joe and I are working on that will also go in. We’re hoping that quite late in the process, in fact after Revcom approval, but again I need to coordinate that with the IEEE. We will be submitting the approved standard (the one that’s passed the ballot) to Revcom in April. That will be considered in the June meeting. Hopefully in the June meeting they will approve the '08 standard. Then, during the fall the standards association would approve that draft on Revcom recommendation and the publication will take place before the end of the year. So, that’s the schedule as I see it currently. Joe Daniels, pipe in if there’s anything you’d like to add in there for us.

Joe D: I appreciate all the work that people did; it has been a success. This is the only standard that I’ve ever worked on which hasn't needed a recirculation ballot. We could potentially have done that also on the previous version, in my recollection there’s only one other standard that I’ve heard of in the EDA world that’s achieved that. What I’m going to do, Andrew is work with Michael Kipness who’s in the IEEE liaison role and have a quick period of phone calls or emails to agree the changes and have IEEE agree how we need to do it if we don’t need a recirculation ballot and then go forward. I suppose the hardest part is no later than April to submit the appropriate set of materials to Revcom, with the minimum touch up work needed on the standard itself. That goes to DAC where the committee meets to say yes, we followed all the correct procedures. They meet June 11^th. The SA board meets the following day, Revcom makes the recommendation, the board says yes or no, and hopefully the week of DAC there will be a formal notification. The formal publication process will start in fall, I’ll do some work in the background and hopefully in that point in time Andrew will also do some work. My belief is that there should be a published standard later this year. December looks very feasible but we’ll know later when we speak to Michael.

Andrew: Thanks. Any other questions or comments about the schedule or the recently concluded ballot?

Joe H.: What in the whole process, other than you and Joe shepherding some stuff through, does the working group need to take in at any given point?

Andrew : That’s my next point Joe ... I feel like you’re looking over my shoulder here, ha ha

Joe D.: Andrew, there is one thing there even though we’re not doing a recirculation ballot. Lets assume that’s true. Those six or seven editorial changes should still get passed by the WG so they can nod their heads informally and say yes.

Andrew: Good point; thanks Joe. Just answering Joe Hupcey's question on what the WG will be working on next. I have this jotted down as P1647-2010, which will be the next revision of the draft. The question on the table that I’d like to put towards those here at the meeting this morning is “Should the working group immediately proceed with work on the 2010 standard, slowly draft the first, considering technical suggestions and changes and soliciting for technical input, but not proceeding at any rapid pace, or should the WG put things on hold for a while and touch base perhaps on a bi-monthly basis and consider the reception to the 2008 standard by producers and consumers in the market place and hold off any parallel development?” Before I pipe in my own opinion I really want to hear a discussion among the WG here. I think, as a note, there are four technical proposals on the table for the next draft that are captured in deferred issues at this point.

Joe H.: My question is actually directed towards this current revision of the standard. Is there anything else the WG needs to do for the 2008 submission that you would anticipate, just out of curiosity, before we start discussing 2010 at all?

Andrew: The answer is, with the exception of the WG looking at the editorial changes, which all of them that I’ve seen are actually like typos, there is no additional work required by the WG.

Joe H.: So you and Joe D shepherd the administrative things along?

Andrew: Yes. We’ll deal with the editing changes and some of the code formatting changes that we want to complete. Sorry for that misunderstanding there Joe. So, back to 2010 and the thoughts for the working group moving forward for the future revision of the standard.

Yaron: I'm not sure if this is related to the long term future ; it more like a requirement to continue operation. Once the standard is out, the license from DASC will expire at that point, because the project authorization is about generating these standards which will be delivered by the end of this year. I think that we should consider again, not in great haste, but eventually, certainly but before the year is out, to put for another project authorization request, another PAR, for the continuing operation of this WG.

Andrew : Okay—good point.

Joe H.: I always thought perhaps incorrectly that the WG could just keep going, but you’re saying we have to apply again? I though it would be automatic but I guess its not.

Mac: The working group continues to support the standard, where if there’s any mis-driven stuff coming down from the approval of the standard then Joe and Andrew kind of handle it to address things. But, to do any extra work beyond correcting typos or paper pushing we require further authorization and that’s what Yaron was saying. So, we’re going to keep going out of momentum for six months just to shepherd this through. Then we should get another PAR to continue to exist.

Joe D.: When the SA board says you are done you have some time to fix things. If new material has to be considered, we need a new PAR. Potentially we have a concurrent new standard and new PAR at the end of the year. So, you would have almost simultaneously a published 2008 standard and potentially if the WG were to do this this fall the PAR being proposed and in place for the 2010 work.

Joe H.: So, you say wait for the current standard to be published before seeking to get a new PAR?

Joe D: That is typically what is done.

Mac: The new PAR will say “revise and extend the existing standard to accommodate new features.” It can be very generic. But it is signing up for another five-year re-enlistment. We do need it to keep working on this.

Andrew: Are there any other comments, questions, recommendations, opinions?

Yaron: Why not list items currently pending?

Andrew: (1) Real number support. Currently supported by one implementation. (2) Temporal coverage proposal from 2004. I had written that at that time and I would like it to be reviewed for consideration. (3) There are some dut_error() variants that have not been captured. They are hopefully small issues which would flush out the current specification of dut_error() to reflect the current Specman implementation and (4) the last issue reflects the fatal() issue. Again, a woefully small change which would be convenient for the e programmer for conveying information back to a caller.

Yaron: There is no thought about adding methodology?

Andrew: None that has been captured at this point as an issue or been discussed within the WG so far as I am aware. When you refer to methodology, you’re talking about trying to better support best-practices today within the language structure?

Yaron: There is some evolution on that front. It contains some significant components as it is. In my mind there is really not much of a question, right? The standards are kept alive or they die. I see no reason not to pursue another PAR and continue the work. We might want to take a breather, we might not want to delve in right away but it seems like the best thing to do.

Mac: I would suggest there’s one thing to do: we solicit for additional donations of ideas and technologies with a six month response time so that we already have a nice list and we should generally be willing to accept others if they’re presented to us in the next time window. Then, write anything we have plus anything we have into the next PAR request and go from there.

Andrew: So you’re suggesting that we use the six month window to wrap up current work, administrative work with the standard and then be prepared in the fall maybe to consider these donations which would have been submitted at that point?

Mac: Yes. You have some donations already. I suggest we finish up the current [PAR] and then issue a call for additional donations with a response date of December 31^st or whatever time the committee thinks is appropriate. If we get nothing, we still have plenty of work, if get additional things then we can consider it in a timely manner.

Joe D: Could be announced around DAC and also announce the solicitation then.

Mac : Right, although it doesn’t necessarily have to be at DAC. Whatever time we announce the final approval you would also, at the same time, say BTW the committee is soliciting for extra donations.

Andrew: We want to convey to the market at DAC that the '08 standard has been passed and is in the process of being published, but work does continue and have evidence thereof which may be as Mac suggested “the technology donation/consultation” in that period of time. I think there is some danger in letting things lie fallow for some time. For instance, getting people back in and the fact that we’re in a WG and the small amount of time, or large depending on each individuals contribution, where we have fallen off their calendar and we need to get back up and running again. Another is the perception in the market place or in the EDA developments world. Is the WG meeting regularly? Are they making progress? Reporting to DASC? Active, etc.? If things look slow it could impact the perception of the language and influence the reception of the language in the verification world.

Joe H: I'm intrigued by the methodology question. For example, sequences that are part of the eRM. I’m wondering Yaron, was your observation about … Cadence is happy to donate anything right? So, my question is more about what makes sense in a standard? Just contributing the methodology we’re running into this thing. The methodology enablers—the language enables eRM itself on top of this—does that get donated as well? Are there other precedents or other standards that not only put in the language enabling constructs but also trying methodology in their standards?

Yaron: I’m not sure. I know that there are different options for standards. One can be where something is a standard for a language and one could be a best practices terminology. If we talk about a full blown methodology it’s more appropriate to go down on its own effort. There are some aspects of usage which come across even with a language standard. Generation would be an example where there is some usage advice in an appendix; that could be a way to go. I don’t have any firm thoughts on this I was just wondering.

Andrew: There are some methodology recommendations in the standard as it stands, right? But, they are not part of the standard proper they’re not part of the “normative” text. They are contextual information to the reader on how to use or where to put particular language constructs, methods how they should be used.

Joe D: It could be an appendix or a supporting P1647.1 supporting use guide. We might not want to go so far today. Even in the original standard what was listed as methodology has changed quite a bit, and we might want to capture that. Mac, what is your experience from the evolution of Verilog?

Mac: A lot of what happened to Verilog in the past year has been merging verification technology and assertion checking into what was originally an HDL standard. I don't know what lessons can be learned for this particular committee. It is good to keep the standard active because nature abhors a vacuum and people either fill it with other ideas or they might think that the standard is dead. There’s been standards that have sat around in DASC for years with nothing happening. Some of the HDL point standards where every year it would be bought up as an example of nothing happening. I don't think six months would make us fall into that trap but I don’t think it hurts to go ahead and ask for donations. We are not doing nothing. We can still meet once a month . Perhaps we should call this to a vote because we are circling at the moment. What is the question?

Andrew: It’s an open question at the moment Mac but it should be stated as a vote and the proposal would be “Should the P1647 WG immediately proceed with work on the 2010 standard or not?”

Joe H: Or digest or something?

Andrew: Or go into the 6 month technology donation period? So, Mark, Charu, Darren, Serrie, what have you got?
Darren: I think there’s plenty you could get going on with what we know already, I mean we’re going to have to get updated documents written for some of the things so there’s no reason why we can’t ask for that under this solicitation period. It’s no different to soliciting from others. It's just soliciting from Cadence.

Charu: I came late to this and have no clear idea here. So, that’s why I don’t want to make any comment on this.

Mark: I guess that I agree with Darren; there is work for the work group so we should get to the new documents.

Andrew: Does anybody want to formulate the question clearly so we can vote on it as Mac suggested?

Joe H.: The question is “Is voting a good idea?” which I guess it is.

Andrew: The question is “Can you formulate the question?”

Joe H.: I'm torn on this. I can see Darren's point, it makes sense to me to continue to meet sometimes and collect solicitations in an informal manner. Things don't have to be captured in formal documents. Flush out things such as floating point that we already know. At the same time we should let this digest to some extent in the market place.

Mac: So can you formulate that as a motion?

Joe H.: The motion would be that the group would immediately try for a PAR for 2010, and work on that PAR or proceed informally until the standard is finally printed?

Andrew: So, should the WG immediately apply for a PAR?

Yaron: May I apply for an amendment? Let's put a date in there because “immediately” is not good. I would amend the motion to be “The group will propose a PAR will solicit a PAR before the end of the year”.

Joe H: The question is, as Darren was just saying, there is stuff we can talk about now, but do we to do this PAR procedure just yet?

Mac: The PAR is Project Authorization Request. For us to do work on behalf of the IEEE, we need to be authorized to do this. The PAR can be research, the definition of life, whatever we like but to operate under the auspices and to be protected by IEEE laws, tax regulations and all that we need to be authorized. So, I would offer that the motion be “Does the group work towards obtaining project authorization for a project to begin sometime in 2008?”

Andrew: So, that would mean the PAR would be re-stating Yaron's amended proposal: Should the group select a PAR then that should be before the end of the year perhaps a little bit earlier? Should the group solicit a PAR before the end of the year, say before August 2008?

Mac: I second that motion so we can call a vote.

Andrew: Any comments or discussion on the motion “Should the group solicit a PAR before August 2008?”

Joe H.: So, that's when we are expecting final IEEE blessing?

Joe D.: No, it's so that the new PAR can be blessed and approved this year. It is independent of the current 2008 one.

Andrew: Any other discussion ? Are there any abstentions or objections to authorizing a Project Authorization Request for the WG by August 2008?

[Silence]

Andrew: Hearing none, that vote has indeed passed. Is there any other discussion under this general status agenda item?

Joe D.: At the next meeting it will be worth discussing if not agreeing on what the PAR is likely to cover. For example, would it just be a series of extensions? Would it be recommended practices? Is it just the solicited donations, new methodology or what? If you’re a study group you’re just trying to solicit donations and figure out where to go. If you want to do a recommended practice, you’re a formal IEEE working group but its definitely a set of recommendations where it’s a language standard or if you want to continue with it, then it’s a full-blown WG just focused on that?

Mac: The study group purpose is to decide whether there is anything worth standardizing and making into a project. We are well beyond that, because, if nothing else, we have to do a reaffirmation of the standard in five years. So, by 2012 to keep the standard going. There is no necessity for a study group. The working group can go ahead and essentially form—open the question of forming a PAR ourselves—because we have an existing group that’s going on ... We should formally ask for donations from Cadence and anyone else who wants to contribute. This is a period that people could be made aware that this is a donation period and people can make donations, then go from there. It is pretty straightforward.

4. Editor

Andrew: Have you had your say Joe?

Joe D: I hope so. After making these changes, I will go away until there is some more work to do, or I am replaced as tasks are reassigned. I’m doing some work in the background to ensure this thing goes through

Andrew: Thanks Joe for all of your work, I personally look forward to working with you and completing the '08 standard.

5. Other Business

Andrew: Please remember to mail rollcall@ieee1647.org to record attendance.

6. Call For Essential Patents

The IEEE-SA updated the patent policy effective April 30, 2007. The details are available in updated slide set at: http://standards.ieee.org/board/pat/pat-slideset.pdf (See “Patent related” from the IEEE 1647 home page, “Slides about IEEE Patent Policy”). If you believe that any patent claims are essential patent claims, please inform the working group. Review the posted slides for more information.

7. Next Meeting

Andrew: April 14th, 9 am PDT.

Joe H: It is the founder of the e language's, Yoav's birthday.

Andrew: I will personally invite him.

8. Adjourn

Mac made a motion to adjourn the meeting.

Darren seconded the motion.

Meeting adjourned at 9:18 am PDT.