W3C home > Mailing lists > Public > public-vision-newstd@w3.org > August 2010

RE: Early draft of new standards proposal

From: Michael Champion <Michael.Champion@microsoft.com>
Date: Wed, 11 Aug 2010 16:40:20 +0000
To: Harry Halpin <hhalpin@w3.org>, Thomas Roessler <tlr@w3.org>
CC: Dominique Hazael-Massieux <dom@w3.org>, Thomas Roessler <tlr@w3.org>, "Ian Jacobs" <ij@w3.org>, "public-vision-newstd@w3.org" <public-vision-newstd@w3.org>
Message-ID: <11A2C786A2D08A4D9E64E213468F4C623E1604F4@TK5EX14MBXC110.redmond.corp.microsoft.com>
> Although I might add that a new process may be necessary if there are not as many "new ideas with substantial impact" 
> coming to the W3C as are needed to keep the W3C as a centre of innovation.

I think this is the key problem we need to address, urgently.  While I agree that we ultimately want to keep the menu of process options down to a minimum, in practice a process document change requires wide consensus across the consortium, and that takes lots of time we don't have.  
I believe the team/Director could experiment with a new process such as the "new idea forum" without having to seek AC consensus.  When we have some evidence on whether it actually addresses the problems, we can address the process question of whether it supplements or replaces XGs, IGs, or both.

My vision of the "new idea forum" is different that www-talk.  I think effective brainstorming of new ideas needs basicially 3 things:
- A focused mailing list, or some mechanism to effectively create topic-specific threads on a general mailing list
- A wiki and/or document repository where more coherent documents can be stored and shared
- A clear IPR policy that informs people about the implications of making proposals and comments.  I suspect the OWF framework is the most appropriate for this kind of thing, but perhaps we could present a menu that includes "None -- you speaks your mind and takes your chances",  the OWF, and the PPs now associated with IGs and XGs.  

Some meta-requirements would include:
- A minimal process overhead that allows anyone to participate as long as they supply an identity to associate with agreement to the IPR policy and to make participants accountable for their behavior (e.g., so we can ban spammers)
- Evangelism about the forum to the target audience.
- Some understanding that successful brainstorming exercises could be published as Notes and the team would facilitate creation of a Rec track charter (perhaps with some minimal level of participation by  W3C members).  Or the result could be submitted to some other body such IETF or OASIS as the community decides.

The governance question is tricky -- do we want to facilitate some tight community with a benevolent dictator decision model to gain traction and credibility under our umbrella?  Or conversely, do we want a wide-open, flame-throwing religious war within or across topic areas?  If not, how do we stop them?  I do not think we can afford to agonize about that now -- if this is explicitly treated as an experiment, the "Director" can change the rules or shut it down as problems become apparent.  At some point, we can bake the lessons into the process document.

-----Original Message-----
From: public-vision-newstd-request@w3.org [mailto:public-vision-newstd-request@w3.org] On Behalf Of Harry Halpin
Sent: Wednesday, August 11, 2010 2:37 AM
To: Thomas Roessler
Cc: Dominique Hazael-Massieux; Thomas Roessler; Ian Jacobs; public-vision-newstd@w3.org
Subject: Re: Early draft of new standards proposal

> On 4 Aug 2010, at 10:59, Dominique Hazael-Massieux wrote:
>> The table is quite useful; it would probably be useful to highlight 
>> what are the things that are still to be defined, since that's the 
>> places where we need concrete input as soon as possible. Also, I'm 
>> not sure if you're interested in feedback on the proposed items for 
>> the new processes or not (that wasn't part of your question); I'm 
>> sending it in any case :)
> I'll throw another dimension in: Are there any of the existing 
> processes that we should throw out? If, for example, we accept the "community group"
> idea, then it's not clear that there's a need for incubators to 
> continue existing.
>> I'm not sure if the so-called "new idea forum" is anything more than 
>> www-talk? If so, and given that most of the lines in the table are 
>> N/A, I'm not sure it's really worth keeping in.
> +1 -- I don't think we should try to introduce a process where the 
> +best
> thing to do is to live without one.

Although I might add that a new process may be necessary if there are not as many "new ideas with substantial impact" coming to the W3C as are needed to keep the W3C as a centre of innovation.

I think obviously IGs should be replaced with/merged with "new idea" forum and XGs should be replaced with/merged with a new "community process".
Having too many process options for folks is actually counter-productive if they are too similar. I'll send another e-mail out on this later.

>> I think the "mostly boilerplate" allusion for the charter of W3C 
>> community groups should be removed; if anything should disappear from 
>> a lightweight charter, it should be the boilerplate. My view is that 
>> a group's charter explains to the world what the group is working on, 
>> and helps the group stay focus on a reduced set of topics; I think 
>> that remains useful no matter the type of the group; but it may be 
>> that the charter should not be a prerequisite to start the work, and 
>> that it in fact be one of the first work items of the group.
> +1 to that as well.  The basic idea could be that some incubator-like
> organism can be put together for a limited time, gets to write the 
> charter, and is only then advertised to the AC.

The issue with charters, as told to me by some of the people we've interviewed in the Social Web XG, is that many people would just rather do some code and set-up a listserv, and eventually draft a spec, rather than write a charter, create a spec, and *then* code. The charter is a barrier to entry for people who are not used to writing them.

However, the problem is not actually chartering, but a lack of communication of people with W3C Team so that they can write a charter together. Also, writing a charter for a WG is currently a pretty big step, often seen as endorsing a particular technology of the W3C. It seems sensible that whatever new process is baked up here makes it easy for coders and folks from outside the W3C to get in touch with a Team member, talk over issues with them, and then let the Team member write a first draft charter *without* necessarily the W3C committing for this to be a new standard.

>> "Approval required": if a chartered group is responsible for 
>> approval, it probably needs to be provided with a set of criteria, 
>> which I guess this task force should at least start drafting; is a 
>> PigML WCG OK? is a WCG on political topics OK? is a WCG on whale 
>> fishing OK? Getting any kind of review in less than one business day 
>> is going to be hard, no matter what. I think that again, there is 
>> probably a need to distinguish various approval/review steps (one is 
>> about getting some basic infrastructure e.g. to discuss a potential 
>> charter which should probably granted very freely and in almost no 
>> time, start working on a concrete spec, etc; another is 
>> reviewing/announcing a charter once it solidifies, gathering IPR 
>> commitments, etc.)

Obviously there should be a review process (and longer than a day will probably be necessary, but hopefully less than a month, but this is a detail), but it should open, transparent, and democratic. For example, there are many technology areas like identity that are outside the expertise of Team (where I think tlr is our only expert in that area).

> The question you're asking boils down to another one: What's the 
> governance model (if any) for these groups?  E.g., do we expect to 
> have groups around competing technologies as a matter of routine?  Do 
> we permit a community group that pursues a technology that competes with a WG?
> What's the expectation for community consensus around charters? What's 
> the expectation for consistency with (some version of) web architecture?
> It'll be important to be fairly explicit about that sort of thing, or 
> at the very least about processes and decision-making.


>> "Consensus-building requirement" is ambiguous: is it inside the 
>> group, or for the group with the rest of the world?
>> This actually opens a different question that probably need to be
>> raised: who is allowed to join a WCG and when? can someone be refused 
>> from joining a group? can someone be ejected from a group?

+1. I think W3C process needs to be clarified here as well *in general*,
such as the case where a Working Chair goes "off the rails" so to speak.
Received on Wednesday, 11 August 2010 16:40:58 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:44:40 UTC