W3C home > Mailing lists > Public > public-w3process@w3.org > August 2014

RE: Overall structure of the W3C

From: Michael Champion (MS OPEN TECH) <Michael.Champion@microsoft.com>
Date: Fri, 1 Aug 2014 21:38:09 +0000
To: Wayne Carr <wayne.carr@linux.intel.com>, "public-w3process@w3.org" <public-w3process@w3.org>
Message-ID: <e0bc585199b843a299ed441ed93aff12@BLUPR03MB488.namprd03.prod.outlook.com>
As Jeff points out, there is no such thing as "W3C" apart from the multihost structure.  So as much as I find it bizarre, incomprehensible, inefficient, etc., I don't see it as a fruitful AB topic.

>>     b. Is the Team's size and makeup appropriate to our current 
>> workload, considering our limited resources?

The harder and more actionable question is whether the W3C team's high overhead  is sustainable.  I agree it's a hard topic for a public list, but we can't reasonably make substantial changes without a good understanding of how non-member stakeholders would react to changes.  I think we should entertain concrete proposals for how to maintain the W3C core value proposition with less Team engagement in the process details, and follow up any rough consensus with an experiment to test an idea's viability. 

Productive area for experimentation might be one of the domains where there is a lot of overlap between the scope and membership of CGs and WGs.  A few that come to mind are

WebVTT http://www.w3.org/community/texttracks/ and TTML http://www.w3.org/2012/02/timed-text-wg-charter.html

http://www.w3.org/community/respimg/ and the HTML WG

http://www.w3.org/community/ortc/ and http://www.w3.org/2011/04/webrtc/ 

It's the last one I know the most about (but what I say has the most risk of being influenced by Day Job Politics, so I'll try to stay away from the details right now). The lesson I take from this for W3C is that expert communities get together in whatever venue they find most congenial and productive, and it will be in a CG (or WHATWG) if the WG doesn't offer a high return on the investment in overhead.  While it seems inefficient to discuss essentially the same technology in different CGs and WGs, maybe we could experiment where instead of a dedicated WG chartered to "own" some space, potentially there could be multiple lightweight CGs in the space.  Whichever CGs can successfully incubate a spec might   then go to an overarching SuperGroup to standardize the spec or reconcile conflicting specs from different CGs.

This approach MIGHT allow overall reductions in team resources, since the team would focus on the well-incubated specs and a few supergroups, and not spent years trying to "grow specs from seed from stony ground" in a large number of WGs.  

BTW, this all my personal musing wearing my AB hat, not anything endorsed by me as Microsoft's AC rep.

-----Original Message-----
From: Wayne Carr [mailto:wayne.carr@linux.intel.com] 
Sent: Friday, August 1, 2014 10:34 AM
To: public-w3process@w3.org
Subject: Re: Overall structure of the W3C

On 2014-08-01 03:27, Charles McCathie Nevile wrote:
> On Fri, 01 Aug 2014 05:02:21 +0200, Nottingham, Mark 
> <mnotting@akamai.com> wrote:
>> On <https://www.w3.org/wiki/AB/2014-2015_Priorities>, the first 
>> bullet in "Overall structure of the W3C" is:
>>> 1. Is the Consortium's current heavy weight structure that was 
>>> created in 1994 still needed now?
>> and Chaals comments: "We don't use the process we had in 1995 or even 
>> in 2005. This question is rhetorically sound but irrelevant."
>> I have to disagree here; this is THE question that the AB should be 
>> addressing. If there's a problem with how the question is phrased, 
>> that's easy enough to fix:
> Yes. That was the problem, and I agree that the AB's big permanent 
> task is to look at whether the W3C is appropriately organised...
>> 1. Is the Consortium's current structure appropriate to the tasks at 
>> hand and the resources available? Specifically:
>>     a. Is the multiple-Host model helpful to the goals of the W3C, or 
>> a hinderance? Are there alternatives?
> My member view is that it is far too focused on the US, with a nod to 
> Europe, meaning the multi-site model is not solving th problem is was 
> meant to solve.
>>     b. Is the Team's size and makeup appropriate to our current 
>> workload, considering our limited resources?
> Good question. This is something that I think belongs in the scope of 
> the AC, rather than a public group.

I'm not getting the focus on team makeup and size.  That seems like micromanaging.  An alternative, that may get at what this one may have been about, could be to ask if there are alternative ways to get more done with limited resources.  There seems to be a move towards this already with the increasing use of Community Groups to do some of what otherwise would be done in more heavyweight WGs (more heavyweight in both process and resources).  CGs could be used more broadly for early spec development and for more experimental specs.  A consideration there would be how to ensure CGs that are tightly connected to W3C are consensus based, since CGs don't have to have charters that ensure that.  The CG charter template was an effort to get at that while leaving CGs lightweight and flexible.

A question could be whether to expand what is already happening in CGs or whether it would be useful to duplicate it in a similar, lighter weight type of WG.

There are already decisions about whether things should go in a CG (one that mirrors a WG or one that mirrors an IG) or an IG or a WG. The AB could look at whether that helps in issues about the impact of the membership model or need for resources.

>>     c. Is the Membership model effective in furthering the goals of 
>> the W3C? What other options are there?
> IMHO, I think it is important. Although it is not entirely 
> representative (and to a certain extent shadows the US-heavy focus), 
> having stakeholders actually holding a financial stake is on balance a 
> Good Thing™ in the effort to ensure that W3C is moving in sensible 
> directions.
> But alternatives may be interesting, and are worth looking at.

Alternatives to membership I'd think would be constrained by the current need to raise a lot of money.  If it was decided some other model would be better than membership, the question would become how to raise a lot of money or how to change the structure not to need a lot of money.

> cheers
>> As a Member, I'd especially like to understand what the multi-Host 
>> model brings to the table; we don't hear much about it, nor the 
>> activities of the "Steering Committee" (see 
>> <http://www.w3.org/Consortium/Agreement/Appendix1-2013.html> section 
>> 3g), which "sets overall policy and provides strategic guidance and 
>> review of the Consortium's activities."
>> Cheers,
>> -- 
>> Mark Nottingham    mnot@akamai.com    https://www.mnot.net/


Received on Friday, 1 August 2014 21:38:38 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:35:11 UTC