Proposed W3C Process 2016 (Call for Review)

Dear AC representative, Chair,
AB, member of the public,

On behalf of the Advisory Board, I am pleased to announce the advancement of the 9 November 2016 version of the Process Document to "Proposed Process Document" status:

See also the diff document showing changes from Process 2015:

Please review the current draft of Process 2016 and indicate whether you endorse the proposal as the operative Process Document, or object to its advancement, via this form:

The deadline for responses is 23:59, Boston time on 9 December 2016,

Comments should be entered into the public issue tracker:

or sent to <> and may be copied to any list you think relevant to the comment.

For Jeff Jaffe, Chair, W3C Advisory Board,
Charles McCathie Nevile, Editor, W3C Process Document,
Steve Zilles, Chair, W3C Technical Report Development Task Force;
Coralie Mercier, Head of W3C Marketing & Communications

The major changes in this document and their rationale are outlined below:

*Renumbered - 5.2.8*
Which becomes 5.2.7 (there was no section 5.2.7 in Process 2015)

*Removed the statement that the W3C doesn't organize conferences.*

*Added a process to make a Recommendation Obsolete and consolidate it with Rescinding a Recommendation - 6.9*

The AB observed that there are some Recommendations that have(mostly) outlived their usefulness, are no longer representing best practices, or have not been widely adopted and are no longer recommended for implementation in new software. This class of Recommendations, called Obsolete Recommendations,  is different from the class of Rescinded Recommendations. Rescinding is an existing process that has an effect on patent licensing commitments (see section 5 of the patent policy). This change only adds obsoleting (and un-obsoleting). A Recommendation that is Obsoleted remains a Recommendation, it still has patent licensing commitments and it can be referenced Normatively, but implementation of that Recommendation is no longer recommended.

Section 6.9 of the Process Document has been changed to add (to the process to Rescind a Recommendation) a specification of the processes to Obsolete and to Restore a Recommendation. The details of the process are similar in each case, but the effect is different. For a few reasons — to streamline the process, because it’s a simple yes/no question (that is, the content of the affected Recommendation, except for the Status section, does not change), and because we would only obsolete when we don’t know of anyone to contact to ask for wide review — Wide Review prior to the AC (and Public) Review is not required or necessary.

Anyone can request one of these actions. If the Working Group that produced the specification is still extant (or exists as a re-chartered group) then that Working Group acts to recommend that the requested action take place. If there is no such Working Group, the TAG acts to do a technical assessment of the requested action. If proceeding is recommended or the AC appeals a rejection, then an AC Review and the Director’s Decision determine the result. See 6.9 for the exact details.

*Changed the voting for AB and TAG elections to Single Transferable Vote - 2.5.2*

The W3C Membership recommended that W3C experiment with different voting mechanisms for TAG and AB elections. After analysis of the 2-year experiment that occurred as a result of that recommendation, the Membership supported the adoption of an Single Transferable Vote tabulation system for TAG and AB elections with the expectation that it will be more representative of the Membership's will.

The text that is in the proposed Process document was designed with the following goals in mind:
 ·The tabulation system description (and choice of specific tabulation system) should be independent of the process document text.
 · The tabulation system should be described independently of specific voting operations (e.g., the forms that members fill out).
 · The tabulation system should be described independently of any software we use to compute results (that is: we should not rely on a single piece of software for implementation).

The Team currently believes that the Meek STV tabulation system is the best fit for the TAG and AB elections. Details on why and how are at:

*Simplified and Rationalized Appeals, so they can occur whether there was dissent or not, and in a broader range of cases – see especially 6.4, 6.6, 6.9, 7, 7.2, 7.3, 10, 10.4, 11*

Toward the end of the process of creating Process 2015, a number of issues related to "appeals" in the W3C process surfaced. At that time, there seemed to be too little time to appropriately address the issues with the care that seemed to be needed. These issues (and ones which have arisen since then) are addressed in the proposed Process 2016.

These changes made the following clarifications:
 A. Which of the three types of appeal is to be used MUST be explicitly identified. The three types are:
 i.   Group Decision Appeal
 ii.  Submission Appeal
 iii. Advisory Committee Appeal
 B. Who can initiate the appeal MUST be identified (whether it is an individual or an AC Representative)
 C. What is being appealed, what "decision" and who (chair, Director, W3C or Team) made it MUST be identified.
 D. There should be a specification of what DOCUMENTATION should accompany each type of appeal. This is specified for a Group Decision Appeal.

Note: Formal Objections are not strictly an "appeal". They are "registered" not "initiated" and they follow the document to which they apply. A separate step, the Group Decision Appeal, that asks the Director to "confirm or deny a decision" (of the group) is the appeal mechanism. Any individual may register a Formal Objection, but only group participants may issue a Group Decision Appeal and if they belong to a Member organization then they must do so through their AC Representative.

Finally, the rules for what decisions are appealable were simplified to be uniform across each class of decisions.

*Clarified the rights and obligations of Member Consortia and their representatives - 2.1.2*

The Problems:
When we introduced the Introductory Industry Membership level [4, 5] we imposed limitations on the rights and privileges of this category of Member. The proposed change eliminates the disagreement between the current terms of an Introductory Industry Member per their Member Agreement and this section of the Process which implies such Members may participate in (all) Working Groups and Interest Groups.


In looking at the way we define the entitlements of Member Organizations that are also a Consortium in nature, there are a couple of issues that need to be addressed.  They arise from the fact that we allow these Members to appoint four (or more) people to represent them within W3C. While we say they are there to represent the Consortium we have been experiencing cases where these designated representatives are in fact representing their own interests.  This opens an IP exposure for W3C because we don't have commitments from their employers, just from the Consortium.  It also offers a "back door" for large corporations to participate without joining themselves.  The proposed changes, in section 2.1.2, attempt to close those loopholes. In addition, the definition of what a Member Consortium is has been tightened to make it clear that a shareholder corporation is not a Member Consortium.

*Clarified the process for continuing work on a specification initially developed under another charter - 5.2.3, 5.2.4, 5.2.6, 6.2.2.*

When the W3C Patent Policy and Process Documents were drafted, some Members may have assumed that work on a W3C Recommendation would be the product of work under a single Working Group charter; instead, Working Drafts often evolve through multiple Working Group charters. The major uncertainty has often been phrased as "When do Working Groups end?", but in fact concerns the situation where a Recommendation is developed under more than one Working Group Charter.  Many specifications take more than one charter period to move from First Public Working Draft to Recommendation. There is a longstanding practice of adopting a Working Draft that was published under a previous charter, and continuing to develop it in a Working Group with a newer charter.

The changes apply to Working Drafts that have had a full exclusion opportunity under a Working Group pursuant to the Patent Policy (i.e., Reference Draft (RD) issued within 90 days of a First Public Working Draft (FPWD) and a Candidate Recommendation (CR) (called Last Call Working Draft (LCWD) in the Patent Policy).

The changes in this draft cover:
 a) A change in the W3C Process Document to clarify how work can continue under a new Working Group charter on a Working Draft that has already had a full exclusion opportunity; and
 b) Suggested improvements in practice to improve the ability to trace the origin of Working Drafts, and their associated Reference Drafts and Candidate Recommendations.

The most relevant text, currently in section 5.2.6 Working Group and Interest Group Charters, is:
“For every Recommendation Track deliverable that continues work on a Working Draft (WD) published under any other Charter (including a predecessor group of the same name), for which there is an existing Reference Draft or Candidate Recommendation, the description of that deliverable in the proposed charter of the adopting Working Group must provide the following information:

 · The title, stable URL, and publication date of the Adopted Working Draft which will serve as the basis for work on the deliverable
 · The title, stable URL, and publication date of the most recent Reference Draft or Candidate Recommendation which triggered an Exclusion Opportunity per the Patent Process
 · The stable URL of the Working Group charter under which the most recent Reference Draft or Candidate Recommendation was published.

These data must be identified in the charter with the labels "Adopted Working Draft", "most recent Reference Draft", "most recent Candidate Recommendation", and "Other Charter", respectively.

The Reference Draft is the latest Working Draft published within 90 days of the First Public Working Draft or if no Public Working Draft has been published within 90 days of the First Public Working Draft it is that First Public Working Draft. It is the specific draft against which exclusions are made, as per section 4.1 of the W3C Patent Policy [PUB33].

The Adopted Working Draft and the most recent Reference Draft or Candidate Recommendation must each be adopted in their entirety and without any modification. The proposed charter must state that the most recent Reference Draft or Candidate Recommendation is deemed to be the Reference Draft or Candidate Recommendation of the Adopted Working Draft, and the date when the Exclusion Opportunity that arose on publishing the First Public Working Draft or Candidate Recommendation began and ended. As per section 4.3 of the W3C Patent Policy [PUB33], this potentially means that exclusions can only be made immediately on joining a Working Group”

*Other changes are in 5.2.3, 5.2.4, 6.2.2*

Note:  Except for Section 3.1 of the Patent Policy, there is no explicit statement in the
Patent Policy that commitments made under the Patent Policy ever expire.

The proposed draft that was circulated for comments on 3 August of this year included the following sentence (twice, once in 5.2.4 and once in section 5.2.6)
“The Director must not issue a call for participation less than 60 days after the beginning of an Advisory Committee Review for a charter that continues work on a document that has had a Reference Draft or Candidate Recommendation published.”

This sentence is not in the proposed Process that is out for AC Review. This 60 day waiting period (waiting to begin a follow-on instance of the Working Group) was requested by the Patent and Standards Interest Group (PSIG). The purpose of the waiting period is to allow companies with patent portfolios adequate time to assess the proposed charter, the documents listed in that charter and to to determine if they have patents they wish to exclude upon joining the follow-on Working Group following the Call for Participation.

Comments received on the August 3 draft suggested that this 60 day period could lead to significant delays in starting the follow-on Working Group, especially if the charter had to be re-Reviewed due to changes made to reflect comments received during the initial Review. This would make the W3C seem less agile and would change the existing practice of allowing a Call for Participation (and starting the follow-on Working Group) as soon as the Review had be successfully completed. It was further noted, that the Review period is 30 days and that there is a normal administrative delay of about 15 days before a Call for Participation with the new charter goes out. That means that there would typically be a 45 day delay before the follow-on Working Group began business. And, in addition, since the W3C tries to be accommodating to its Members, if one were to say that required more time to do a patent review prior to starting the follow-on Working Group, then it is very likely that the delay would be granted. Note that, because the list of documents (and their status) is in the charter, no new First Public Working Drafts or Candidate Recommendations would be published during this period so that the Review has a consistent set of documents.

After presenting the above information, the question of whether or not to include the 60 day waiting period was posed to the AC Representatives attending the AC Meeting on the Thursday of TPAC 2016. No AC representative wanted ot include the waiting period and approximately 15 AC Representatives wanted it not included. An additional, unspecified number of AC Representatives expressed no opinion. For this and the above reasons, the waiting period sentence was deleted from the draft that has been sent out for review.

*Reverted an earlier change that made making Editorial Changes unnecessarily more difficult. 6.7.2*

A change was made in Process 2014 that required AC Review of a Proposed Edited Recommendation, a Recommendation that did not have substantive changes, only editorial ones. This was a change from the prior Process 2005. This change was seen as making the W3C less agile (a comment received on the 3 August proposed Process 2016) and discussion agreed with this assertion and noted that because the Working Group doing the edits had to make the recommendation to re-publish the (Edited) Recommendation without dissent, what other group could better assess that the changes were only editorial. Thus the requirement for an AC Review was seen as only delaying things and not likely to change what is published. The AC Review, in this case, has, therefore, been eliminated in the proposed Process 2016.

Coralie Mercier  -  W3C Marketing & Communications - +336 4322 0001

Received on Thursday, 10 November 2016 10:20:39 UTC