W3C

A Peaceful Ear

Jeffrey Zeldman has written a weblog entry An angry fix about Björn Hörmann‘s message on his reasons for leaving the group doing the development of W3C validators. He made a few points in his message which will be certainly discussed by the Web communities in the following days.

Jeffrey starts his message with:

Some of the best minds working in web standards have been quietly or loudly abandoning the W3C. Björn Hoehrmann is the latest.

Björn is neither leaving W3C nor being abandoned by it, quite the opposite. He has, for example, just accepted a role as invited expert within the CSS working group. He says it in the conclusion of his message: Well, that’s it for me then; more time to work on the CSS and WebAPI specifications. He’s saying that it seems for him difficult to be able to continue to develop the Markup validator (on the HTML side) bearing with issues in HTML specifications which are not answered by the HTML WG.

Jeffrey touches a few issues about the possible levels of participation to W3C:

Alas, the organization appears unconcerned with our needs and uninterested in tapping our experience and insights. It remains a closed, a one-way system. Like old-fashioned pre-cable TV advertising. Not like the web.

Writing that the W3C is a “one way system” is not a very accurate characterization. Centralized, perhaps, and it is true that W3C working groups prefer receiving feedback rather than hunting for it in thousands of individual weblogs, but the feedback system W3C provides could hardly be called closed:

  • Comments are opened on the first Working Draft which are published by any working groups.
  • New Working Groups often also starts their activity with a requirement documents which collect the needs and what the technology should be and people can send comments to those as well.
  • … and all of above is generally made on public mailing-lists

Not many other organizations have such a level of openness. Of course it could be made better, and the W3C as a whole is trying hard to listen to everyone – thus creating and animating such informal discussion and information places as this QA blog.

Jeffrey is then praising for coherent specifications and interoperability.

We require coherent specifications based on our and our users’ actual needs. Upcoming accessibility and markup specifications fail on both counts. We require validation tools that work and are kept up to date. Instead, tools are still broken years after problems are reported.

Yes we all do. There are a few issues to solve for achieving this which I hope we will find an answer. Small closed communities are a lot easier to manage. The problem is often how do you reconcile conflicting user needs in an open large organization like W3C.

Any suggestions are welcome.

10 thoughts on “A Peaceful Ear

  1. Imagine an architect who designs a house after conferring with committee members financed by a corporation.

    He hasn’t spoken to the people who are supposed to be buying the house. He hasn’t asked them what they want. He and a committee have figured out what they think a good consists of.

    Next the architect posts the blueprints on a bulletin board in a grocery store. He hopes the potential homeowners will notice the posting and comment on whether or not they like the house he has designed for them.

    What would work better?

    It would work better if the architect talked to the homeowners, found out if they had children and needed rooms for those children, asked if they owned a car and needed a garage, found out what kind of architecture they like, and so on.

    You are right that Björn quit a group, not the W3C. I stand corrected.

    My point is that, unless it enjoys creating specifications for their own sake, the W3C should make an effort to consult with web designers and developers at the outset.

    WCAG and XHTML are two examples of emerging specs created by brilliant and well-meaning people but with no apparent connection to the actual work our community does.

    I understand why you are defending the system, but I wish you were listening more and defending less.

  2. Jeffrey, we (QA Boys) hear you.

    The architecture metaphor is a very good one, except that you did not pick the right building. If technologies were buildings designed by architects, then web standards are not individual house (that would be a personal Weblog), we are talking about a community space like an hospital for example. And how does one have an influence in how a hospital is designed and built? Such projects are designed by committee, with representatives of different communities (W3C Working Groups), and regular town hall meetings (Working Drafts) requesting comments from the public (feedback comments on mailing-lists).

    In such a metaphor, you would agree family having discussions about the new planned hospital in their living room or at the corner of a neighbourhood street (personal weblog, forum) hardly constitutes feedback, and the architect is not able to hear those. He, and the city hall, are listening, however, and the family’s comments will be more than welcome at the town hall. (mailing list of the technology being developed)

    We agree that “W3C should listen the Web designers and Web developers”. Just like architects and town halls, W3C has a formalized process for public feedback. If this process is not good enough, then it should be fixed; if it was breached, then that must be fixed too. We agree.

    But are you complaining about that process, we can’t know. Have you, personally, sent feedback about HTML specifications that were not treated properly? Apparently, as of today, not. So, are you complaining about a breach in the process, or the feedback process itself, or something else? To our question “how can we better listen to Web designers and Web developers?”, your only answer is “yes, you should listen”.

    Could you develop that? we are listening… Tell us how do you think we should collect comments of people in a practical and effective way?

  3. Björn writes:

    Now, This is another comment, where I discuss this in more detail. […] So, does the group have a good explanation for that? I’ll tell you, once they bother responding.

    Isn’t this a big enough hint of where the communication process fails in the first place? It’s about feedback – both directions.

  4. Robert,

    No argument about that at all. Björn made very interesting points in his emails which have to be taken into accounts on te way his comments have been handled by the HTML WG.

    In general, someone raises an issue on a mailing list (Björn did), then the chair and/or members of the WG decide to record the issue in an issue tracker for discussion. Once discussed the Group replies to the commenter by either a disagreement, a partial or a complete approval. (consensus building with conflicting issues from multiple sources.)

    What Bjoerns stressed in his email is that some issues, when made at the right place, have not been answered, that is a problem! And when some WGs are making mistakes or are unable to deal with those because of too much work, we (W3C) fail somehow and need to fix that ASAP!

  5. I am confident that the W3C are aware of relevant issues (more so than the rest of us) and will continue to steer the web’s development in the right direction. Group members come and go, but the goals remain.

  6. zeldman writes “the W3C should make an effort to consult with web designers and developers at the outset”. Yes, of course. And I think we do try, to some extent. Steven Pemberton goes to Web Design conferences and such.

    The community is so large that talking to everybody seems overwhelming. But I’m interested in suggestions.

    Meanwhile, the analogy of the architect and the home buyer seems off by a bit: in the design of HTML, there are zillions of people in the position of the home buyer, no?

  7. Unlike Mike Whitehurst, I am confident that the W3C is almost completely unaware aware of relevant issues (less so than the rest of us) and will continue to steer the Web’s development in the wrong direction. If you think that’s too broad, I would be happy to restrict it to accessibility and markup.

    At this point we need Karl to stop reiterating “Did you try using our feedback mechanism?” when what we are saying is “Your feedback mechanism doesn’t work for real Web developers and has been rejected. Yes, you really do have to search thousands of Weblogs, and tools made by people who built upon W3C specs because they were tired of waiting will make that easy for you.” I mean, come on, I search thousands of Weblogs.

  8. I may be way off base but it seems to me that working on specifications isn’t all that different than working on a large software development project. In fact, writing the specifications in some cases is one step removed from software development so why not treat specifications as software? Instead of using a mailing list to gather feedback use something like Trac and make it open to the public. Each section and subsection is immediately available for review as it’s created. If a particular section or sub-section has a problem, someone opens a ticket.

  9. Scott,

    This is already the case for most WGs. See

    ITS WG and XML Core during their review period use Bugzilla opened to the public to record comment AND Mailing list.

    Joe Clark, just before seemed to say that Web developers didn’t know how to use mailing lists to send comments, I don’t see how it will make easier for them to use a bug reporting system. Mailing-Lists offer a good compromise for reporting and then chairs can use the features of Bug tracking system to keep an history of the records.

    As for tracking weblogs, *I* do often look at Weblogs, but I can _only_ read French and English with enough understanding to be able to report to a group or to invite people to report on the Mailing List. Plus the fact that WGs have already hard time to get the job done on issues they are already receiving if they have to hunt new ones, that will make it more difficult. I’m pretty sure Joe Clark knows how to track Chinese, Japanese, Thai, German, Spanish, Russian weblogs (and many other languages), but I would like to know his dirty secret. :)

    What I could imagine and that would be more scalable is that Web standards organizations like WASP, Openweb, W3QC, etc in their own countries collect the information and report it to the appropriate WGs. That would be a real community process.

Comments are closed.