Meeting 30 Minutes

Teleconference, May 8th, 2007

A quorum of 3 out of 3 voting members present. 5 non-voting members also present. Attendance and voter status indicated on the roster web page.

-----------------------------------------------------------

Attendees
* Voting members
  o Yaron Kashai
  o Joe Hupcey
  o Andrew Piziali
* Non-voting members
  o Joe Daniels
  o Serrie Chapman
  o Darren Galpin
  o Henry Von Bank
  o Mark Strickland

1.      Call to Order

·         Reminder: Mail a message to “rollcall@ieee1647.org”

2.      Approval of Meeting 29 Minutes

·         Andy solicited motion to approve the minutes

·         Serrie made the motion

·         Yaron seconded the motion

·         No objections, motion passed

3.      Status Update

Andy : we have three primary subjects to talk about, the first is voting membership the second is the overall status of the review of the clauses and lastly we'll talk about the specific status of the donations and where they are at this point in time.

·        Voting membership

Andy : So, if you've visited the online roster recently in the past month or so you'll notice that currently there are only three voting members of this working group and the reasons vary as to According to my records, there are currently only three voting members in the working group.  The reasons vary either because of lacks of meeting attendance, because people that have been attending meetings are not members of the DASC.  So, in order to maintain those voting privileges, you have to meet those requirements, as I was mentioning to Joe Hupcey before the meeting was actually called to order this morning.  A substitute for meeting attendance is a discussion of a topic recorded in the posted minutes and that topic could be discussed or should be discussed in the e-mail reflector.  It occurred to me that you can review the minutes and discuss the subject with the e-mail reflector to do so.   So, I guess that if you want to have some influence on the language above and beyond technically contributing, which will become more important later this year, I would encourage you certainly to send a note to the reflector to let those know that are not here **** to maintain your voting status.

Any discussion on that subject?

Yaron : Andy, may I make a proposal?

Andy : Sure.

Yaron : The reflector is a very broad audience so whatever you send to it you’re going to have this generic and polite flavor to it.  But many of the people registered there have never attended a meeting.  I think that we may want to have a more …subject .. contain … mainly…????****.  So, what I suggest we do is go down the list and look at people who have been attending and inquire with them whether they would, I mean directly, we need to take into account that having a voting membership of 3 would be looked upon as insufficient support by the IEE.

Andy: right

Yaron: Its not enough participation. So, beyond going to the reflector and giving a generic heads up there I would actually follow up with individuals that have been attending in the past.  Encourage people like for instance Matan who was very active to go ahead and register with DASC.

Andy: That’s a great suggestion and in fact I intended to mail individual messages to those who have been quasi active or, for example, active in the review process but perhaps haven’t been attending meeting .. or for some other reason they haven’t met the voting requirements.  Likewise your advice to mail a polite message to the reflector as a whole, given its broad audience is well taken.  Any comments on the current voting membership or ideas on how to encourage people to take a half hour or so out of their time each month to attend this meeting?

Mark:  I’m still kind of curious over why we want this under DASC at all? I mean the SystemVerilog stuff isn’t under there its under an IEEE committee.

Andy: That’s true – alright Yaron maybe you have some background on that as I actually don’t know have we ended up being sponsored by DASC?

Yaron: right – there has to be a sponsor body for SystemVerilog its CAG or “CAG” and **** its  DAC??  CAG is like a last resort sponsor body that works on the core entity balloting, so not for the stuff that we’re doing which is individual ballots.  In general , EDA standards have been sponsored by DASC for as long as anyone can remember and it is the more appropriate body to be sponsoring these kind of standards on top of that, they are the platform on which we operate so they have been responsible for the PAR.  We can’t just decide to pull out or something without re-issuing a new PAR and finding a new sponsor.

Mark : Is it different when you ..

Yaron : For one thing its hugely expensive and in my mind its  hugely political, even more so than DASC if you can imagine that.  Mark I’m curious, what your concern is behind DASC versus another sponsor body… what are your thoughts there?

Mark : I can see being a member of IEEE per se, but DASC seems to be just an EDA body right, they’re talking about new power initiatives and things like that.  I don’t care – I prefer to be a thing that was just reviewing the standards like yours.

Yaron: So, you would argue that this language standard doesn’t belong in the … or should be sponsored by the design automation standards committee?

Mark: Yeah – I guess I just don’t understand why that is.  I can understand before its an IEEE standard and you would need some kind of sponsor but I would have thought that once it was IEEE, you’d just have to be an IEEE member and you could vote on improvements to the standards.

Yaron: Sadly this is not hoe things are set up.  IEEE has an SA?****these are things that are in place and we have zero influence on.

Mark: So, why is it I don’t have to be a member of CAG to vote on SystemVerilog?

Yaron: In fact you can’t be a member of anything to vote on SystemVerilog right ? – You need to be a corporate member of IEEE to do that kind of voting.  So, individuals cannot take part in the vote.

Mark: But I know that individuals do – not that I do but

Yaron: Yeah – the corporations are represented by individuals

Mark : Yeah – but there are consultants on there they’re not part of a corporation

Yaron : They have an entity membership somehow, so, they work through these …trust me they are both more expensive and complicated than what we’re doing.

Joe Hupcey : A brief history SystemVerilog is, in fact, a driving ****candidate by the CAG and the **** and the reality is the CAG is set up for entity based balloting so, you are correct Markus? Mark?

Mark : Mark.

Joe H : Mark, so you are correct that people can participate in the meetings but in terms of the formal votes and especially the balloting its only done in terms of entity.  It’s the way that entire things set up, its structured deliberately that way.  so, for an individual ballot like 1647 was and originally 1364 was, a system which is, you need to be a member of the sponsoring organization and you need to be an SA  member.  So, for the corporate based, the entity based ballot stuff you need to be an SA member and you need to be a CAG member, and by the way the CAG membership project fee is 3750 a year whether you’re an individual or a member of academia – theres some trade off to go with that.  It appears like everyone’s participating the same and they are relative to influencing work group meetings and decisions but in terms of the final vote and especially the balloting stuff, its only done at an entity level.  so, Cadence gets one vote, Microsoft, if its participating in that ballot gets one vote etc etc that’s the way it works.

Andy: Okay – thanks for the verification .  and DASC, as I recall, annual membership is something like $40 a year is that correct?

Joe H: and by the way mark it used to be that if you were a DASC member that was all you needed to be then IEEE said that you needed you needed to be an IEEE member and now you have to be an SA member and potentially if you want to participate in an individual standard then you’ll pay hundreds of dollars but if you want to participate in the other standards then you’ll pay thousands of dollars.

?: you have to participate if you want to vote

Andy: Any more discussion ion the voting requirements?

Mark : Yeah I have another , maybe an observation  or suggestion.  It seems the boards and things are pretty active – I thought maybe there’d be some co-credit, if you will, for people who are particularly active participants with posting and that sort of thing, but, for whatever reason they just can’t make the phone call.  It seems like they should be given credit for that activity.  You’ve kind of addressed this but is there some obvious criteria where you’re going to make this, postings or something – this should count.

Andy: Actually just a single posting per minutes posted per meeting is sufficient – I don’t have it in front of me the precise by-laws phrase but I reviewed it recently and had some discussions with Serrie and with Darren and I think I may have forwarded that to them but essentially if you review the minutes of a meeting and post some discussion onetime about one subject on that meeting then that does credit you for attending that.

Joe H : Okay – that’s a great way to re-summarize it.  I know you’ve summarized it before but those last three sentences have **** - maybe that’s something that people may be more aware of.

Andy: Okay I’ll mention that in a more summarized form in the message I send to the past active folks – the folks that were active in reviews and postings but haven’t met the other requirements.  Anybody else?

• Review status

Andy : Okay well the status update is currently we have about 90 issues that have been opened for five of the six clauses that have been circulated.

The one that has not seen any of the issues posted is the sequences clause which we posted the first draft last Wednesday. If anybody has a chance to review that document I encourage you to do so.  Joe did a great Job of taking a large initial document and reducing it down to about 12 pages.  so, now its just what sequences are all about and all the syntax and semantics behind it.  The other top level or **** status concerns bullet point here that although I asked for volunteers to share the task force to review these open issues, to advise the editor on how to satisfy the concerns – we haven’t had yet any volunteers.  This is not a large amount of work, again we’re talking 90 issues which were evenly distributed – its about 15 issues per clause right?  They would really just need to be looked at and compared to the existing clause with the existence with one or two other people that might work with you due to an interest in that particular clause, suggest changes, additions or deletions to the clause to the editor.  so, I’ll encourage you to consider that and I’ll certainly volunteer people in the near future if I don’t have volunteers.

I’ve gone through about half of those so far in assigning those to the various task forces that need to be completed so I’ve got about 40 left to go.  In the process of doing so I’ve updated the subject lines of most of those to include the clause they deal with so we don’t have to poke around and figure out what the writer is talking about.  Any questions on the review status.

Joe H. : Andy, what are the nature of the reviews that just kind of confirm that whatever things people expect are … behavior today is captured in the revised standard or is it more like you have to sit down with pencil and paper and write out the code examples and make sure it really Joe H. : What’s the sort of the nature that has to be done??

Joe D: Sure.  Mark asked the same question actually, so let me address that .. So, except for a very small number of issues which in fact do ask for an example and sometimes in fact we do put examples in the LRM.  Most of the work involved is to read the issue as described, read each section of the clause which is discussed,  and then make a determination as to whether such an ambiguity exists and needs to be clarified, whether there's a technical error and that error needs to be corrected through an additional change to the LRM, the clause that’s going into the LRM or whether perhaps somebody is suggesting a language change that is beyond the scope of what this clause aims to do and adds to the language and in fact needs to be rejected.  So, it varies a little bit but most of the issues, from what I’ve seen so far are straightforward, either a grammatical error or an ambiguity or theres conflict perhaps between something n one section and another section.  I haven’t seen many issues that I’ve looked through so far that addressed a technical concern where they addressed for example an error, although Matan has caught several of those.  Joe Daniels I would welcome your comment on this.

Joe D. : The idea behind this is that each reviewer Joe,  will take a look at it with different eyes so for example there’s actually <<name>>***who's one of the technical writers who so far as the technical group took over a few more relative to suggesting eyes down on grammar.  Generally speaking people will say technical stuff and if they’ve been able to actually run the code – say Matan finds something he runs the code against it and then he can really walk in and say “its wrong because of this reason here”.  Ideally that would be done if it’s a technical issue but more than anything else is a confirmation of .. to review that the right  materials there, its presented in a fashion that’s reasonable, I can follow it and heres some smaller things that need to get either expanded or fixed.

Andy: So, does that answer your question Joe H?

Joe H : It does. I’m trying to determine to what extent, I used to be technical but my atrophy kind can, you know, look at some of these things and sat oh yeah – I can see .. I can pitch in here or not. 

Joe D: I think that for a lot of that in fact you could and the intent is that these task forces be composed of 2 or 3 people that you could work with.  For example one of the engineers, 1 of the other e users or within cadence for example.

Joe H: Let me suggest why I’m unlikely to because at one point 3 years ago I was a fresh set of eyes and I guarantee if you haven’t looked at these in a while you’ll pick up things that you’ll probably mostly think upon as this doesn’t really make sense or you’ll see a reference for some hole where we’re anticipating that someone familiar to the material will holler **** it and I’ll become familiar with the material and I’d like someone else to take another look at it as well. 

Andy: Even in the early iterations between Joe D and I, looking at the material that’s going into a draft clause we see things that are not seen by another and that dialog has been quite valuable.

Mark: I understand that for the purpose of creating issues.  How does that work for reducing the number of issues or filtering some of them out?

Andy: Lets Postulate an example Mark.  In the process of reviewing the sequence document I’m a member of the working group and lets ignore the straightforward issues such as there’s an error or  there’s an option missing or a keywords missed out or something like that.  Let’s say that a person says it would be nice if sequences supported this additional capability so it ought to have this additional keyword and this additional option which influences how sequences operate, so in other words it’s a language extension proposal.  Is that the scenario you wanted to – for example address?  A lot of the others are like black and white – like grammatical fixes, ambiguities that need to be clarified and things like that.

Mark: Yes – say with grammatical errors, the committee might just say ‘yep’ it’s a real issue it should be fixed and it goes off to the technical writing team to fix it.

Andy: That’s right so that person simply signs off on status change of the issue

Mark:  Okay and then the next issue is … I need more details here and that would go off, not to the technical writing team but I suppose to some engineering team?

Andy : Well actually, probably the original source of the clause, in all likelihood a cadence R&D person will look at the document and that document may or may not, in fact, have the clarification.  If it does, the tech writer, Joe Daniels, can go and try and pull that information in to clarify it.  If it’s in fact missing in the original documents then we would work with the author of the originals document to find out what the clarification should be.

Mark: So, the output of this team is exactly that – say – this looks like a grammatical mistake, this looks like a missing detail this look like something that’s outside the scope of what we’re doing …

Andy: Right and each one of those resolved would be a recommendation as to what the status of the issue should be changed to and it would go back to the tech writer for the change that needs to be done.  If its closed or its dismissed because we’re not going to consider such a change for the 08 LRM ..

Mark : And, what would be the downside of just having the tech writer do this?

Andy: Well, Joe what do you think the downside would be ?

Joe H.: I’m not sure what I add to the process if say for example I was working on one of these to say if its grammatical or not.  Any tech writer could identify that. If, lets say and issue says that we’re missing some detail and I say yes I got it confused too, that would be what the tech writer would do.

Lets say that we’ve got a comma wrong here etc etc and it goes in a queue for me to fix. A more complex on eis something appears to be missing here, someone else within the task force or somewhere else agrees that something’s missing here, I probably need an additional person to provide that information to me and that’s the time that **** because someone that’s more familiar with the material, especially **** tool level can come up with that quicker than I can.  And the third case is when its clearly incorrect, there was never anything in the original source documentation then that definitely has to come from someone else.  Then if the work group says its out of scope then its deferred until the next revision past 2008 when we’ll address this or we’ll try to address this through some other form.

Mark: Okay so say somebody else found some issue saying I don’t understand this and I look at it and say I do understand it, does that means that its not an issue?  I would think it was still an issue because the first person had a problem with it.

Joe: What I’m hoping in that case is that within the issue tracking database the dialog is covered, I think it’s the history mechanism – I need you Andy to talk about this as I’ve never known how the **** works, and finally a resolution to the issue which either comes to me or is closed, because its an interpretation issue which doesn’t need an update to the standard.

Mark: So, there’s a means of tracking a dialog between the person who opened the issue and a task force member who’s reviewing the issue?  They can come to an agreement that a particular change to the clause has satisfied the original author’s objection.  I’m still really stuck on what the reviewers value add is here because if the originator had some confusion it seems like some kind of change should be made because this person represents some percentage of the population who are going to read this.  And if some change should be made what does the reviewer bring to the party that the tech writer couldn’t do as far as working out with this person what would clarify it?

Andy: I think its primarily, mark just a technical dialog that needs to take place between the technical person who opened the issue and a second person familiar with the language that acts as nothing other than another person to bounce this idea off of.  Am I missing something? In other words if you and I were looking at some e code and we’re trying to understand it, or you wrote it and I think it behave differently than you believe how it should behave then that dialog alone helps clarify behavior.  If it does clarify behavior then we would go back, perhaps, and talk about why did we interpret the code in a different fashion.  I think the same thing is true in the clauses of the LRM itself.

Andy: The idea is that  there's going to be some concerns **** that’s how standards are made.  Its probably inappropriate for individual members to bring their concerns to the editor and have the editor serve as some sort of broker/Arbiter.  The idea is that whoever is reviewing these issues is doing that in a mindset of this proper perspective of trying to find something that will satisfy everybody.  I have to say that everything that is flagged as an issue is a real issue.  There’s some miscomprehensions on the part of the people introducing the issue but the idea is not that there’s going to be one person’s core judgment versus another, its going to be more about trying to reach some compromise that everybody’s happy with.  Is that okay Mark?

Mark : I still kind of see the reviewer as the middle man just filtering information between the person who filed the issue and the original author or tech writer.

Andy: The person who takes charge of a specific clause is a person who’s actually driving the creation of this compromise.

Mark : So, that person, they own the issue and its resolution versus the original author of the issue who is on fact proposing a change or flagging a problem that needs to be addressed?

Andy: So, to summarize, Joe Daniels can obviously pick up issues of grammar and fix that but if its some semantically error in the e example or there’s something that’s omitted. Joe or myself might not necessarily pick it up.  Sequences, this too, there’s not a paragraph on stuff like that -  this too right? Is that the sort of thing you’re looking for?  Somebody would be the owner of that and maybe write the paragraph and say hey you’ve missed this part of the sequences too , they also do this and they slice bread as well and heres the sliced bread paragraph and Joe would go off and look at it.  And that person would be the owner right and they would check that Joe had implemented it.

Andy: Joe would also be looking at the grammar, any English language related issues Joe D takes care of.  Sometimes Andy.  Anything that’s e-related technical members of the working groups deal with that.  So, I guess that Marks happy with trying it so we’ll try to organize some small task force to look through the open issues.  Theres no other discussion on that status then I want to quickly touch base on the specific status on the donations.

Specific 1647-2008 donations

Name Spaces (annex C)

Encapsulation

Method Ports

Messaging

Constant Fields

Andy: I'm in the process of classifying those, as I said earlier and assigning them to each particular task force.  I have a local spreadsheet and I can capture that information in the issue database itself as an assignment.

Sequences has been posted for review last Wednesday and I’m currently looking for people to walk through those pages and put anything that you find in terms of wanting to discuss in terms of sequences.  Just out of curiosity has anybody on the call yet reviewed that clause at all?

Darren: Not yet.

Andy:  The last is the reflection API and  Joe Daniels is in the process of writing the initial draft. We’re trying to get a copy of the original source which the original document was written in – we’ve got the PDF but Joe Daniels’ job would be a little bit easier if he could pull across the precise font and things from the original.  I contacted Matan but I haven’t heard back from him, if you want to contact him yourself or give me his phone number I’ll contact him myself when he’s bright eyed and bushy tailed in the morning.

Yaron: We tried to look up that document – apparently it was some kind of internal document that was initially created in word and then was translated into FrameMaker – we never found the original.

Andy : Oh – neither the Word or the FrameMaker?

Yaron: Yeah we had the Word document which wasn’t exactly the same.  The conclusion was that it was more trouble than its worth.  The guy who wrote it was Avi Ashar, that would be the most knowledgeable person beyond Matan.

Andy: I think Joe you have the Word document?

Joe : yeah, Yaron the reason I thought there might be a more recent FrameMaker document is because the PDF that came from Matan have comments that appear to be from a technical writer that appears to be commenting.  So, I thought that maybe a FrameMaker copy exists somewhere also and if that exists then lets start with that otherwise I’ll use the PDF and start from there .

Yaron : I would guess that it wouldn’t hurt to send an email to Jay right? And ask him if he has anything recent.  But last time I tried it didn’t amount to anything.

Joe : All we got was that Word version

Yaron: I’ll send a note to Avi – who goes by Jay actually, and see if he has a copy in FrameMaker.  I’ve got that action item.

Andy: Any questions on the specifics of the clauses and where they are in their review or process

4.    Editor

Joe: I just want to see if Andy knows …..do you use the reflection process as my view is that this is something that should be **** by itself and 2 I think it’s a smaller number of people who are using this and therefore  **** not in the clause part….

Darren: Could you speak up a bit please Joe?

Serrie: Sorry I struggle to minute your input mostly but this week its really quiet.

Joe : IS THIS BETTER?

Serrie : Yes; much better. Thank you.:o)

Joe: I would like to find out if people are currently using the reflection facility because my belief is it should stand by itself it really does almost need a completely independent standard obviously though its related to the primary.  Because of that I would prefer to put it in as an appendix because I really believe that this is going t o be used at best by a very small set of people in the world.

?: Joe, what is the status of the PLI for Verilog do you know? Is it in an annex?

Joe: The PLI for Verilog is a set of clauses.

?: It is a set of clauses?

Joe: Yes.

Andy : It sounds like the use of the PLI, VCPI and all that maybe… my experience is, as I told Joe,  that I know of two or three people in the world that have used the reflection facility among the e-users.  The one guy I was hoping to coax into reviewing the document doesn’t want to look at it so unfortunately I don’t have that option.  But this question is for you, Yaron as well as Joe Hupcey and maybe Mark, how widely used is the refection facility and how should it be positioned within the LRM?

?: if theres only 3 users in the world I wouldn’t say it’s a runaway hit

Yaron : There are 3 users and maybe 5 if you include R&D and I know R&D are like a law unto themselves.  So, maybe its something that  **** uses it for *** work so ..maybe we can ask her to take a look.

Andy: Another thing that comes to mind from when I was talking to Joe Daniels about this would be in the Unix world of system calls, there is a small number of calls that are documented along with all the other system calls.  They’re only used by the people who write debuggers so theres a ptrace system call, for example, and the ptrace call and its brethren and they’re not called out separately into some special area because only debugger users are using them.  Other than that, they’re simply not widely referenced that section of chapter two of the Unix man pages.  The reflection facility appears to be very similar in the way it’s a way to introspect and therefore its not widely used and is a kind of special purpose set of facilities.

Joe: It has to be captured somewhere and can be off to the side of **** widely used and even though she doesn’t attend the meetings **** we do use this facility so she can at least take a look at it for us.

Mark: We do use it extensively here at CISCO so I don’t know what it means to me as a user whether its an annexe or it’s a clause.  So long as its in the document I don’t really care.

Joe: I need to hear someone outside of Cadence using it.

Andy: The difference Mark would be that as an annex its kind of like an appendix and you might well have a hard time finding it.  2 – it may not seem like … I think most of our annexes Joe today are not normative – I’ve forgotten what the other term was …

Joe : This is what I need to hear so Andy, unless you send me an e-mail today saying otherwise it will be a clause.  That’s fine no problem.

Andy: So, Mark you said it is widely used within Cisco?

Mark; Yeah – its used by knowledgeable people, its in the core eVC code that we share

Andy : Okay. TI was the other company I was familiar with that was making use of it.  Any thing else you want to touch base on ? Joe?

Joe: No. Just so folks are aware that hopefully sometime next week or later on this month I will give the draft of the reflection API.  It will probably be about 20 pages and it will be a very target set of reviewers that will be looking at that.  However it will need a most thorough review because it is the newest material.

Andy: Another thing I might mention just while Joe has the floor is that he and I are going to meet at DAC on Wednesday June 6^th to review some of the current status of the LRM issues, that’s going to be in person.  So, if you’re going to be at DAC let us know and we’ll make sure we coordinate face to face with everybody who’s going to be there.

5 .     Other Business

Andy : The only thing I had was that I did double check and update the dial-in instructions for toll-free numbers.  I checked US, UK, Israel and India and they worked as published on there. There was a slight change to the dial-in so if you have any questions then take a look at the minutes page on the website.  I didn’t fly to the UK and Israel and India to do that. The magic of Skype lets you place calls virtually from any of those countries.  Any other other business?

6.       Call for Essential Patents

Please review the mandatory slides posted 

The IEEE-SA updated the patent policy effective April 30 

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.

An essential patent is a patent that would be infringed if it were not licensed or transferred to the IEEE in order to file it legally as standard.  So, if people leave with any claims that are essential patent claims they should meet with the work group and take a look at the posted slides for more information on the subject.

I’ve had a look through the PDF, but not being completely familiar with the policy nothing obvious jumped out at me.  I don’t know if Yaron has even looked at it? – I think they may have just tightened up some of the existing procedures.

Yaron : Yeah; I looked at it briefly and I couldn’t spot any differences its not the first time, they often do this.  Once every couple of years they go through and they revise everything – to me it looks pretty much the same, I think I did right by not becoming a lawyer as I can’t tell the difference… 

7.      Next Meeting

Andy: Tentatively set for June 11th at 9 am PDT (1600 UTC)

Is anybody aware of any holidays that fall on top of that ?

Serrie: There’s non in the UK.

Andy: Yaron are you aware of any Israeli Holidays?

Yaron : Whats the date again

Andy : June 11^th

?: Monday right after DAC

Yaron: No I don’t think so 

8.      Adjourn

Andy solicited motion to adjourn.

 Darren made a motion to adjourn the meeting.

Serrie seconded.

Meeting adjourned.