Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: [802SEC] Should all IEEE 802 drafts coming for sale be a CLEAN file or should they be offered as they come (in the recirculation case, with changes marked)????




Roger:

Creation of the initial draft is a always a challenge and it goes easier on some projects than others.  In some of our projects there is a high level of trust and early work is informal and largely by concensus.  

Larger groups or groups with strong technical divisions have to be run more formally during the selection of technical alternatives (e.g., which line code do we use, which of the 26 PMD proposals do we standardize).  Typically after a few drafts these groups get less formal only resorting to formal votes on specific items when there is controversy.

In 802.3 the initial draft is usually based on slide ware.  The slide ware proposal must be accepted by 75% for inclusion within the draft (either in the initial draft or to fill one of the gaping holes).

The first draft is an attempt of the editor(s) to faithfully turn the slideware into text, tables and figures.  Usually, the draft is accepted without controversy, but occasionally, the editor doesn't recognize something as being controversial and that portion of the draft may need to be removed (if 75% of the group doesn't want it in it is out).  Basically everything left in the hacked up D1.0 has either explicitly received 75% approval in its text/figure/table form or as part of a blanket approval of the draft (typically a motion to incorporate changes agreed at the meeting and produce a new draft).

Editors are encouraged to use editor's notes (not part of the draft and to be removed prior to publication) when creating new material.  Usually, slideware leaves something unspecified, ambiguous or incomplete.  The editor uses judgement to either:
1.  flag that something is missing with an editors note;
2.  if possible put in a TBD (e.g., perhaps the proposal only specified a max value that will go in a table, then the editor would make a min that is required TBD);
3.  flag text he/she created in producing the first draft but believed to be implicit in the proposal;
4.  or if likely to be controversial may include possible text within the editor's note.  

Once we have an initial draft accepted by 75% vote, all additions/deletions/changes must also be by 75% vote.  The changes may be created using concensus rather than formal votes on individual items (e.g., does anyone object to ...), but production of a new draft for release to the committee is always done with a 75% vote.  (As indicated before, this is usually something of the form: "accept the changes made to Dx.y and produce Dx.y+1".)

Similarly, in comment resolution we do a lot by concensus, only resorting to formal votes on details when there is controversy.  We will include formal votes when rejecting a negative ballot comment where the commenter is not satisfied with the committee response.

In summary, we stay informal when the group and requirements of process allow.  We get formal when there is controversy, a batch of informal work needs ratification, or process dictates.

--Bob Grow

-----Original Message-----
From: Roger B. Marks [mailto:r.b.marks@ieee.org]
Sent: Tuesday, April 15, 2003 2:10 PM
To: Grow, Bob
Cc: stds-802-sec@ieee.org
Subject: RE: [802SEC] Should all IEEE 802 drafts coming for sale be a
CLEAN file or should they be offered as they come (in the recirculation
case, with changes marked)????


Bob,

Thanks for this explanation. I appreciate the constructive dialog; it 
gives me a better perspective on your process.

It looks like there are a lot of similarities between the 802.3 and 
802.16 processes. This isn't surprising, since I have learned a lot 
from the generous advice of 802.3 regulars. In particular, the 802.16 
Task Group Review and Working Group Review processes arose only 
because the 802.3 people explained the advantages of the Task Force 
Review.

It looks like our approach is quite similar from the point a Working 
Group Letter Ballot opens. Beforehand, the main difference is when we 
define a document as a draft and to what extent we distribute it.

There is one question I've been wondering about but never remembered 
to ask anyone in 802.3: how do you use the 75% approval rule? Does 
adoption of a document as D1.0 require 75% approval? At that point, 
does any change to it require 75% approval? Or are things handled 
more loosely very early in the process?

Roger


>Roger:
>
>Herein lies a difference between working groups.  We don't worry 
>about the semantics of draft / working paper.  We set the bar at 
>technical completeness as judged by the Working Group.  I strongly 
>support the position that prior sale of the flaky, incomplete, etc. 
>drafts is an invitation to disaster.  Very early drafts are 
>available to anyone interested enough to become a participant in the 
>project to gain some of the context Geoff talked about.
>
>In 802.3, once we have a PAR, we start building the draft.  We 
>include the IEEE copyright information and keep it in a password 
>protected web location.  Initially, we are adopting proposals and 
>turning slide ware into prose, tables and figures.  The document at 
>this point often (but not always) has obvious gaping holes. 
>Documents at this stage are typically identified as D1.x.  We may 
>produce documents that are not distributed to the Task Force (i.e., 
>editor's drafts for use by the editorial team). 
>
>On our large projects we will often do a formal task force review. 
>This is run somewhat like ballot (put out the draft and ask for 
>formal comments, teach the TF how to use the ballot tools, etc.), 
>except the entire document is always open for comment.  Sometimes we 
>bump the draft number to D2.x though we might still be at D1.x (as 
>802.3ah is currently).
>
>When the Task Force believes it is technically complete they request 
>WG ballot.  We will again bump the major draft number (e.g., D3.x) 
>and start a formal ballot with the scope of the ballot narrowing 
>during recirculation on the changes per IEEE balloting rules.  When 
>the WG ballot is announced, we tell IEEE to offer the draft for 
>sale.  Each recirculation ballot we offer the new draft for sale. 
>During recirculation, we typically produce both diffs from the 
>previous draft and a clean (no change bars) version of the draft.
>
>When we enter Sponsor ballot, we again typically bump the major 
>draft number (e.g., D4.x) following the same procedure as described 
>for WG ballot.  As a consciousness raising exercise (reinforcing the 
>narrowing scope of a recirculation), we often only distribute the 
>pages that actually have changes (e.g., for the current 802.3af 
>recirculation we only supplied the 34 pages with substantive 
>changes, though the complete clean version is still available).
>
>--Bob Grow
>
>
>
>-----Original Message-----
>From: Roger B. Marks [mailto:r.b.marks@ieee.org]
>Sent: Tuesday, April 15, 2003 12:31 PM
>To: Howard Frazier
>Cc: stds-802-sec@ieee.org
>Subject: Re: [802SEC] Should all IEEE 802 drafts coming for sale be a
>CLEAN file or should they be offered as they come (in the recirculation
>case, with changes marked)????
>
>
>
>Howard,
>
>Let's talk practically and see where our traditions agree and where
>they don't. Our normal approach is that the motion to adopt a first
>draft and the motion to open a WG Letter Ballot go hand in hand. We
>password-protect the draft and put it up for sale.
>
>If our document is not worth the pixels it's displayed in, we don't
>call it a draft. We might circulate it (under some kind of Call for
>Comments, which we might call a Task Group Review or Working Group
>Review), but we are careful to NOT call it a draft. We normally call
>it a "Working Document", and we label it carefully that way.
>
>So, regarding the sale of drafts, I think that semantics play a role
>here. From my perspective, we don't label a document as Draft 1 until
>we think it's ready for WG Letter Ballot and are not embarrassed to
>have IEEE sell it.
>
>If a Working Document isn't ready to be a draft, it's still a
>"committee document" and, as I read CS rules, we must make it
>available to "all interested persons." So we do.
>
>Roger
>
>
>
>At 11:48 AM -0700 03/04/15, Howard Frazier wrote:
>>   > With this rule available, I do not believe that there is any doubt for
>>>   Angela to strongly push for streamlining the process to make ALL draft
>>>   standards available.
>>
>>
>>I would be vehemently opposed to any such policy, and I do
>>not believe that the CS rules require us to make rough, ragged,
>>early, incomplete, inaccurate, erroneous, half-baked,
>>non-sensical, premature, flaky, not-worth-the-pixels-they're
>>displayed-in, drafts available for sale.
>>
>>We have a duty as members of a professional society to produce
>>professional quality work.  Our early attempts at creating
>>a draft standard may represent our best efforts at the time,
>>but they clearly do not represent anything close to the final
>>completed work.  We do not want to disseminate false information,
>>or set false expectations.  We are already grappling with the
>>problem of claims of conformance to draft documents, and this
>>problem would only get worse if all of our early work was
>>disseminated to the public.
>>
>>For these reasons, I strongly support the policy of making
>>drafts available only after they have been issued in the form
>>of a WG ballot.  This should be the norm.  I have consulted
>>with some members of the IEEE-SA staff, and this is their
>>current understanding of our policy, and they think it is
>>sensible. Exceptions can be made on a case-by-case basis.
>>
>>Howard Frazier
>>
>>Hayes, Vic (Vic) wrote:
>>
>>>Roger,
>>>
>>>
>>>Thanks for finding the rule (at the Computer Society) I was looking
>>>for but could not find at the SA site. .
>>>
>>>
>>>With this rule available, I do not believe that there is any doubt
>>>for Angela to strongly push for streamlining the process to make
>>>ALL draft standards available.
>>>
>>>
>>>I would like to encourage all WG chairs to ensure that the draft is
>>>for sale at the time it would also be available to the members.
>>>
>>>
>>>Regards
>>>
>>>
>>>---------------
>>>Vic Hayes
>>>Agere Systems Nederland B.V., formerly Lucent Technologies
>>>Zadelstede 1-10
>>>3431 JZ  Nieuwegein, the Netherlands
>>>Phone: +31 30 609 7528 (Time Zone UTC + 1, + 2 during daylight saving time)
>>>FAX: +31 30 609 7556
>>>e-mail: vichayes@agere.com
>>>
>>>      -----Original Message-----
>>>      From: Roger B. Marks [mailto:r.b.marks@ieee.org]
>>>      Sent: Tuesday, April 15, 2003 6:01 PM
>>>      To: Hayes, Vic (Vic)
>>>      Cc: stds-802-sec@ieee.org
>>>      Subject: RE: [802SEC] Should all IEEE 802 drafts coming for sale be
>>>      a CLEAN file or should they be offered as they come (in the
>>>      recirculation case, with changes marked)????
>>>
>>>      Vic,
>>>
>>>
>>>      I thoroughly agree with your emphasis on the principles of openness.
>>>
>>>
>>>      In my view, the rules that 802 needs to follow on this are actually
>>>      quite simple. They come from the Policies and Procedures of the IEEE
>>>      Computer Society
>>>
>>>      Standards Activities Board
>>>     <http://www.computer.org/standards/ORIENT/p&ptoc.htm>:
>>>
>>>>      4.3 Document Availability
>>>
>>>>      All interested persons shall be permitted to obtain all committee
>>>>      documents, including draft standards prior to approval by the IEEESB.
>>>
>>>
>>>      IEEE 802.16 has always followed this policy. We request that our
>  >>     drafts be made available for sale by IEEE. If, for whatever reason,
>>>      an interested party cannot purchase a draft from IEEE, then we
>>>      provide it directly.
>>>
>>>
>>>      Roger
>>>
>>>
>>>
>>>      At 5:54 AM -0400 03/04/15, Hayes, Vic (Vic) wrote:
>>>
>>>>      Tony,
>>>>
>>>>      I am not arguing against the payment issue. In the paper era, it
>>>>      was obvious that the copying needed to be paid for. Now, it is the
>>>>      organizations view of whether the copyright needs to be translated
>>>>      into an income factor or whether the developers want to pay.
>>>>
>>>>      In the documentation I could only find a section in the Standards
>>>>      Companion that is in line with my definition. The model sponsor
>>>>      rules are more in line with your definition.
>>>>
>>>>      Quote from Standards Companion:
>>>>      Openness is also a principle that applies throughout standards
>>>>      development. It means ensuring that everyone has access to the
>>>>      process. This is accomplished by making sure that all materially
>>>>      interested and affected parties can participate in your standards
>>>>      development group, and seeing that the results of your
>>>>      deliberations are publicly available. The latter is usually
>>>>      achieved by having readily available minutes of meetings.
>>>>
>>>>      The purpose of all this is to avoid the appearance of collusion,
>>>>      or seeming to obstruct anyone from participating. All IEEE working
>>>>      group meetings are open, and anyone may attend if interested. This
>>>>      principle must be employed for every official IEEE meeting. Any
>>>>      person has a right to attend and contribute to IEEE standards
>>>>      meetings.
>>>>
>>>>      Openness also provides protection against antitrust situations.
>>>>      Since standards are so broadly used and often carry the weight of
>>>>      law, it is important to allow all parties to participate and be
>>>>      heard to avoid a situation that would imply that any company or
>>>>      individual was restricted from speaking.
>>>>
>>>>      Both of these principles should be considered from the very start
>>>>      of your standards process. They are vital to the formation of your
>>>>      working group and the creation of your PAR.
>>>>
>>>>      Quote from Model Sponsor rules:
>>>>      The Secretary shall record and have published minutes of each
>>>>      meeting. [The Treasurer shall maintain a budget and shall control
>>>>      all funds into and out of the sponsor's bank account.]
>>>>      and
>>>>      4.1 Voting Membership
>>>>
>>>>      Voting Membership in the Sponsor shall be in accordance with the
>>>>      procedures of the entity that established the Sponsor, or, in the
>>>>      case of a TC with P&P, in accordance with those procedures. In the
>>>>      absence of such procedures, voting membership is open to any
>>>>      materially interested individual who notifies the IEEE Standards
>>>>      Department of his/her interest and provides and maintains contact
>>>>      information, and conforms to the committee rules for attendance
>>>>      and balloting.
>>>>
>>>>      I still feel that all drafts need to be available to the public,
>>>>      whether for free or for payment
>>>
>>>>
>>>
>>>>      Regards
>>>
>>>>
>>>>      ---------------
>>>>      Vic Hayes
>>>>      Agere Systems Nederland B.V., formerly Lucent Technologies
>>>>      Zadelstede 1-10
>>>>      3431 JZ  Nieuwegein, the Netherlands
>>>>      Phone: +31 30 609 7528 (Time Zone UTC + 1, + 2 during daylight
>>>>      saving time)
>>>>      FAX: +31 30 609 7556
>>>>      e-mail: vichayes@agere.com
>>>>
>>>>
>>>>
>>>>      -----Original Message-----
>>>>      From: Tony Jeffree [mailto:tony@jeffree.co.uk]
>>>>      Sent: Tuesday, April 15, 2003 11:09 AM
>>>>      To: Hayes, Vic (Vic)
>>>>      Cc: Grow, Bob; a.ortiz@ieee.org; stds-802-sec@ieee.org
>>>>      Subject: RE: [802SEC] Should all IEEE 802 drafts coming for sale be a
>>>
>>>>      CLEA N file or should they be offered as they come (in the
>>>>      recirculation
>>>>      case, with changes marked)????
>>>>
>>>>
>>>>      Vic -
>>>>
>>>>      All depends on how you define "openness". Taking your line of
>  >>>     argument to
>>>>      its logical conclusion, to be truly "open", there would be no obstacle
>>>>      whatever (including financial obstacles) to free & open access to
>>>>      our work,
>>>>      and so all drafts and published standards should be available to
>>>>      all for
>>>>      free. This is the position that I hold personally; however, it clearly
>>>>      isn't the position that the IEEE holds. I suspect that the working
>>>>      definition of "openness" for the IEEE standards process is much more
>>>>      limited, and is along the lines that anyone who wishes to do so can
>>>>      participate in the work, subject to the membership rules of the
>>>>      committee
>>>>      concerned, and anyone that wishes to read drafts and standards
>>>>      that are
>>>>      made available during the progress of that work can do so, subject to
>>>>      payment of any fees that may be due for the privilege.
>>>>
>>>>      To my knowledge, the decision as to when a draft should be made
>>>>      available
>>>>      for sale has always rested with the working group concerned, and
>>>>      is made
>>>>      when the draft has reached a reasonable level of stability
>>>>      (whatever that
>>>>      might mean).
>>>>
>>>>      Regards,
>>>>      Tony
>>>>
>>>>
>>>>
>>>>      At 04:01 15/04/2003 -0400, Hayes, Vic (Vic) wrote:
>>>>
>>>>      >Bob and Angela, SEC members,
>>>>      >
>>>>      >Because the IEEE-SA does have the requirement to be an "Open"
>>>>      Committee, I
>>>>      >would interpret the question "which drafts are available for
>>>>      sale" to be
>>>>      >answered as "all drafts, even change page instruction as well as
>>>>      versions
>>>>      >with change bars".
>>>>      >
>>>>      >As to Bob's indication that they only make drafts available "once
>>>>      we have
>>>>      >entered WG ballot", I would like to state that they are violating
>>>>      the rules
>>>>      >for openness.
>>>>      >
>>>>      >Regards
>>>>      >
>>>>      >---------------
>>>>      >Vic Hayes
>>>>      >Agere Systems Nederland B.V., formerly Lucent Technologies
>>>>      >Zadelstede 1-10
>>>>      >3431 JZ  Nieuwegein, the Netherlands
>>>>      >Phone: +31 30 609 7528 (Time Zone UTC + 1, + 2 during daylight
>>>>      saving time)
>>>>      >FAX: +31 30 609 7556
>>>>      >e-mail: vichayes@agere.com
>>>>      >
>>>>      >
>>>>      >
>>>>      >-----Original Message-----
>>>>      >From: Grow, Bob [mailto:bob.grow@intel.com]
>>>>      >Sent: Monday, April 14, 2003 9:27 PM
>>>>      >To: a.ortiz@ieee.org; stds-802-sec@ieee.org
>>>>      >Subject: RE: [802SEC] Should all IEEE 802 drafts coming for sale be a
>>>>      >CLEAN file or should they be offered as they come (in the
>>>>      recirculation
>>>>      >case, with changes marked)????
>>>>      >
>>>>      >
>>>>      >
>>>>      >Angela:
>>>>      >
>>>>      >It would be great to have an automatic process, but I am not
>>>>      clear on one
>>>>      >issue.  There is no consistent policy on when drafts are made
>>>>      available for
>>>>      >public sale.  In the case of 802.3, we make drafts available once
>>>>      we have
>>>>      >entered WG ballot.  In this case we do not upload drafts to the
>>>>      ballot
>>>>      >center.
>>>>      >
>>>>      >During reciruclation ballots, we might only distribute change
>>>>      pages for the
>>>>      >ballot.  (For example the upload for the current P802.3af/D4.3
>>>>      recirculation
>>>>      >ballot included change pages only (about a fourth of the complete
>>>>      draft).
>>>>      >
>>>>      >I believe a clean version is the appropriate version for
>>>>sale.     This is also
>>>>      >the only consistent thing we do throughout the entire ballot process.
>>>>      >Because of FrameMaker's limitated diff capabilities, we may
>>>>      change the way
>>>>      >we produce the change bar version depending on the change
>>>>      volume.  Because
>>>>      >the upload isn't the clean version, and it isn't necessarily
>>>>      complete, an
>>>>      >automatic process will include staff picking up the complete
>>>>      clean version
>>>>      >of the draft from the WG private pages.  Some questions need to
>>>>      be answered
>>>>      >for the process to be both comprehensive and automatic.
>  >>>     >
>>>>      >1.  How does staff learn of first public availability of a
>>>>      project draft?
>>>>      >2.  How will staff learn of WG ballots or new drafts prior to sponsor
>>>>      >ballot?
>>>>      >3.  Do all WGs produce and post clean versions of documents for every
>>>>      >recirculation?
>>>>      >4.  Do all WGs announce the URL, username and password for the
>>>>      complete
>>>>      >clean draft on each ballot announcement?
>>>>      >
>>>>      >I support your efforts to make this process automatic, but I will be
>>>>      >concerned if it doesn't also support sale of drafts prior to
>>>>      sponsor ballot.
>>>>      >
>>>>      >
>>>>      >I also think it is important that we be able to invoke this automatic
>>>>      >process without uploading the complete clean draft.  Our voters
>>>>      are able to
>>>
>>>>      >work with pointers to the draft, staff should be equally willing
>>>>      to work
>>>>      >with the pointer (URL, username and password).
>>>>      >
>>>>      >--Bob Grow
>>>>      >
>>>>      >-----Original Message-----
>>>>      >From: a.ortiz@ieee.org [mailto:a.ortiz@ieee.org]
>>>>      >Sent: Monday, April 14, 2003 11:49 AM
>>>>      >To: stds-802-sec@ieee.org
>>>>      >Subject: [802SEC] Should all IEEE 802 drafts coming for sale be a
>>>>      CLEAN
>>>>      >file or should they be offered as they come (in the recirculation
>>>>      case,
>>>>      >with changes marked)????
>>>>      >
>>>>      >
>>>>      >
>>>>      >Hello All:
>>>>      >
>>>>      >In our efforts to keep improving the process to make IEEE-802 drafts
>>>>      >available for sale, there are some things that need clarification.
>>>>      >Therefore, I will like to raise the following question:
>>>>      >
>>>>      >I understand from Jerry Walker that we do not need to confirm
>>>>      with the WGC
>>>>      >any longer, if the draft will be made available for sale, but
>>>>      instead, this
>>>>      >will be a default process, meaning that every time a new or
>>>>      revised draft
>>>>      >comes,  we will make these drafts available for sale.
>>>>      >
>>>>      >With that in mind, I would like to get input from all of you as
>>>>      to which is
>>>>      >the right thing to do in this case.  Hence, please let me know if the
>>>>      >drafts we will make available for sale, are to be **as they
>>>>      come**  (with
>>>>      >the changes marked)  when it comes to recirculations, or if we
>>>>      should make
>>>>      >*only clean drafts* (without changes marked* available for sale.
>>>>      >
>>>>      >Please let us know as we are streamlining this process, of making
>>>>      IEEE-802
>>>>      >drafts available for sale in a timely manner, especially since
>>>>      this process
>>>>      >is so important for all of us, especially for our customers.
>>>>      >
>>>>      >Please keep in mind that the prompt input from every WGC,
>>>>      regarding drafts
>>>>      >coming for recirculations, is needed and very much appreciated.
>>>>      >
>>>>      >Regards,
>>>>      >
>>>>      >Angela Ortiz
>>>>      >Program Manager - Technical Program Development
>>>>      >__________________________
>>>>      >IEEE Standards, 445 Hoes Lane,
>>>>      >Piscataway, NJ  08855-1331 USA
>>>>      >Telephone: 1732-562-3809  ><  Fax: 1732-562-1571
>>>>      >E-m:  a.ortiz@ieee.org   ><   standards.ieee.org
>>>>      >
>>>>      >FOSTERING TECHNOLOGICAL INNOVATION
>>>>
>>>>      Regards,
>>>>      Tony