W3C home > Mailing lists > Public > public-w3process@w3.org > August 2016

Re: W3C Process 2016: Charter extensions and 60 days publication blackouts

From: David Singer <singer@mac.com>
Date: Mon, 29 Aug 2016 16:54:43 -0700
Cc: Michael Champion <Michael.Champion@microsoft.com>, Philippe Le Hegaret <plh@w3.org>, "public-w3process@w3.org" <public-w3process@w3.org>
Message-id: <3706C1B3-FF71-4004-884A-97F7CA5A9330@mac.com>
To: Carine Bournez <carine@w3.org>

> On Aug 29, 2016, at 13:00 , Carine Bournez <carine@w3.org> wrote:
> On Mon, Aug 29, 2016 at 05:48:03PM +0000, Michael Champion wrote:
>>> I still think that mandating explicit "adoption" of all specs when rechartering without
>>> a WG (name) change is too complex in this draft of process2016 and introduces
>>> unnecessary delays and burden (list all of a WG's specs in the charter!)
>> The W3C Process tries to strike a balance between not imposing ???unnecessary delays and burden??? on one hand and not leaving ambiguities that could result in royalty requests or patent litigation on the other.  This topic has been very extensively discussed in PSIG and the current proposal is the result. The Process 2016 language mainly tries to align different interpretations of the patent policy with recent practice when rechartering long-running groups, in the hope of tightening up ambiguities that some find worrisome.  
> Having the process fixing ambiguities in the Patent Policy is very far from ideal. Fixing
> the PP FAQ would have been much simpler, so I think it's not just about ambiguities.
> About 10 years ago, PSIG and others were advising against closing a WG, in fear of 
> losing associated commitments. Now the proposal is completely different, dissociating
> the specs from the groups, while commitments are made wrt WG lifetime + CfPs.

There are also PP FAQ updates that have been discussed, but (a) they are the prerogative of PSIG and (b) can be made at any time. This is the ‘matching’ process change.

> Concretely, the 60 days without FPWD is not sufficient if we have "Reference Drafts" 
> that can be published less than 90 days after FPWD, so for consistency, there should be
> a second condition forbidding to update a FPWD published less than 90 days before 
> the start of the review. OR (I prefer that, actually) we use "FPWD" instead of 
> Reference Draft in the wording. That second option matches the current implementation
> (Ref Draft = FPWD, as long as there's no WG change).

I think there are two questions being mixed in here, and it’s worth teasing them apart.

a) Can I write a charter that clearly delineates which documents are ‘under the patent policy’ as the group forms? Clearly, if documents go ‘under the policy’ AFTER the charter was written, the answer is no. Hence the restriction on doing transitions AFTER the charter text is written and being reviewed.  Do we all agree that those are things that this group will be working on and responsible for ‘ab initio’?

b) For anyone joining this newly chartered group at its beginning, they might wish to file an exclusion on such drafts. If they were members of the originating group, they are aware of it and have been tracking it, but there may be new members. Because the list of drafts is in the charter, they can prepare any exclusion they need to file if given enough time; for various complex reasons, we chose 60 days as the minimum interval for the newcomers to be allowed to think. Otherwise newcomers might be forced to delay their join until after the group ahs formed and started work.

>> I don???t suppose that the current proposal is the best we can do, but ???le mieux est l'ennemi du bien???.
> OK, if it really covers all the scenarios, but I don't think it does with the current 
> wording. Transitioning could be particularly tricky in some cases (e.g. drafts
> published by more than 1 WG that are not scheduled to be rechartered simultaneously,
> and that is just 1 example, there are probably other cases). It's always better to
> anticipate than discover afterwards that we have not sufficiently tested the new rules.

I completely agree.  Perhaps there are some questions of ‘joint deliverables’ we should explore.

>> -----Original Message-----
>> From: Carine Bournez <carine@w3.org>
>> Date: Monday, August 29, 2016 at 12:23 AM
>> To: David Singer <singer@mac.com>
>> Cc: Philippe Le Hegaret <plh@w3.org>, "public-w3process@w3.org" <public-w3process@w3.org>
>> Subject: Re: W3C Process 2016: Charter extensions and 60 days publication  blackouts
>> Resent-From: <public-w3process@w3.org>
>> Resent-Date: Monday, August 29, 2016 at 12:23 AM
>>    On Thu, Aug 25, 2016 at 01:50:32PM -0400, David Singer wrote:
>>    [PLH]
>>>> The proposed Process 2016 introduces the following:
>>>> [[
>>>> Transition requests to First Public Working Draft or Candidate Recommendation will not normally be approved while a Working Group's charter is undergoing, or awaiting a Director's decision on, an Advisory Committee Review, until the Director issues a Call for Participation for the Working Group.
>>>> ]]
>>>> https://dvcs.w3.org/hg/AB/raw-file/cfef536bff0d/cover.html
>>>> Depending on the interpretation, this addition may be either overly restrictive (1) or doesn't have effects (2).
>>    [DS]
>>> Not exactly. I think it???s trying to deal with the problem that the new charter can???t identify the drafts that are adopted that are ???under the patent policy??? if they are published and become such AFTER the charter is frozen and sent for review.
>>    The current proposed text does not solve the question, IMHO. It uses the term Reference
>>    Draft that is usually a FPWD, so 60 days does not solve anything. If we wanted to 
>>    have a frozen list of specs that had or are under an exclusion opportunity, it should
>>    cover the period during which in theory the Ref Draft can change, so 90 days, not 60.
>>    Important note: we currently don't implement a different Ref Draft when a draft is 
>>    published in the first 90 days of a FPWD exclusion: In 2006, an issue was raised,
>>    explaining that changing the Ref Draft during the first 90 days of the exclusion 
>>    period makes those 90 days - which are btw the opportunity to leave without commitment -
>>    useless if the Ref Draft is a moving target. so since 2006, Ref Draft = FPWD, except in 
>>    the case of adoption by another WG.
>>    I still think that mandating explicit "adoption" of all specs when rechartering without
>>    a WG (name) change is too complex in this draft of process2016 and introduces 
>>    unnecessary delays and burden (list all of a WG's specs in the charter!)
>>    The grace period already covers something close to the 60 days. We currently implement
>>    a 7-day delay for exclusion upon joining during the WG's life, we could extend that 
>>    a bit if necessary, without crafting complex rechartering rules and delays.
>>    -- 
>>    Carine Bournez /// W3C Europe
> -- 
> Carine Bournez /// W3C Europe

Dave Singer

Received on Monday, 29 August 2016 23:55:17 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 August 2016 23:55:18 UTC