W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 2003

Why Consensus? ...and how to raise issues

From: Matt May <mcmay@w3.org>
Date: Wed, 17 Dec 2003 07:10:25 -0800
Message-Id: <24617802-30A3-11D8-A7E5-000393B628BC@w3.org>
To: WAI-GL <w3c-wai-gl@w3.org>

I want to attempt to clarify what exactly the consensus process is, why 
it exists, and why it's a good thing. As you read this, note that I am 
a W3C/WAI employee, but have no special status in the working group, 
and am merely trying to get us back to the tasks at hand.

Nobody is held back from participating in this working group, full 
stop. Not participants who don't happen to be in good standing, not 
people who aren't on the list, nobody. The WCAG WG is probably the most 
open working group in the consortium, and I say this having been an 
invited expert myself long before joining W3C.

The discussion here on good standing seems to be taking the angle that 
good standing is some way to keep people out of the process. In 
actuality, it is a way of drawing interested parties in. After clearing 
the invited-expert hurdle, anyone, irrespective of organizational 
affiliation, can become a participant in good standing. The concept of 
"good standing" in the consensus process helps prevent less-interested 
parties from scuttling the work of the group.

A scenario: individuals A through W get together to solve a problem, 
and are about to arrive at an agreement. Each individual in this group 
under the consensus process has the right to stop the group and require 
their issues to be addressed.

Individual X has just joined the group. Not knowing how a certain 
decision was arrived at, or having contributed to the process to date, 
X does not have the right to block consensus. However, if X provides 
value to the group and becomes knowledgeable about the topic, the 
consensus process then establishes X's right to cause issues to be 

Individual Y once joined the group, has rarely appeared or interacted 
with the group, but appears for the call for consensus. Y's issues may 
already have been resolved. Y may not be aware of the current state of 
work. Or worse, Y may be trying intentionally to destroy the work 
created to date. For these reasons, Y is considered by process not to 
be able to block consensus.

"Good standing" in the consensus process is designed to keep people 
actively involved in work that benefits the group, and giving each 
participant the power to effect change where it most matters to her or 
him. That power has to come with some kind of check, and for the W3C 
that is in the form of regular attendance and participation. (There are 
other means of breaking a move to block something, but they are 
designed for situations where a compromise cannot be reached, and the 
Director ensures that there is otherwise a genuine consensus among the 

So, who are you if you bear the unfortunate moniker of "bad standing"? 
You're the same as anyone else, at least in the WCAG WG. You can join 
the conference calls, attend the face-to-face meetings, read and post 
to the mailing list. The only thing you don't get to do, under the 
consensus process on which W3C is based, is stop the train. We seek 
consensus from not only the working group, but the approval of the 
other working groups in the WAI domain, then proof that the document is 
viable in the wild (Candidate Recommendation), then the consensus of 
the W3C membership (Proposed Recommendation), and _then_ the approval 
of the Director of W3C before the document becomes an official 

We do not blithely ignore the world around us in this process. Wendy 
regularly receives reviews and new issues from outside the working 
group, brings those issues to the working group, and proposes changes 
to the document based on outside comments. And people in the working 
group who are not "in good standing" have their issues addressed with 
no discrimination as to their standing.

So, then. That's consensus in a nutshell. I'm hoping people are now 
asking, what can I do to get my issues addressed? (Please tell me 
you're asking this. Seriously. I'm not even supposed to be here today. 
I'm on jury duty, and I'm writing this at night while my wife is 
watching the 14-hour Two Towers edition.)

First, go to the meetings, if you can. W3C puts a high priority on 
active participation, and regular teleconferences are to date the most 
effective means for pushing things forward. That's why it's (almost) a 
requirement for participation in good standing. It is very important to 
the process, in that we can ask whether a given item has consensus and 
have a near-immediate answer. For the techniques task force, it's 
important to keep coordinated on what everyone is doing. It is also 
helpful to attend because whether or not the group thought something 
was a good idea or not often doesn't come through well in meeting 

Even if you cannot go to the meetings, read the drafts and comment on 
them regularly. Send emails to the w3c-wai-gl and/or 
public-comments-wcag20@w3.org lists, pointing out the draft you're 
commenting on, the problem you have with it, and if at all possible, 
potential solutions. If you have expertise in a given area, take an 
action item to solve the issues you know how to deal with. If you take 
an action item, produce a deliverable, or tell the chairs or staff 
contact that you can't.

Read what consensus means to the W3C, and how issues are addressed[1].

Note that the WCAG 2.0 draft points to an issue-tracking list[2] which 
contains hundreds of open issues with the document. Make sure you are 
suggesting something new, and if not, add your comments to any existing 
bugs against the document.

Raise issues clearly. Start a new thread. State the problem you see 
with the document, citing examples if possible, and what you think 
would resolve things. Raising a concern in one sentence halfway down a 
message on an unrelated thread is not likely to result in the group's 
immediate consideration.

Hope this helps.


[1] http://www.w3.org/2003/06/Process-20030618/policies.html#Consensus
Received on Wednesday, 17 December 2003 10:10:17 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:46 UTC