I’ve been meaning to write a little bit more about this for some time, but was finally inspired to do so when I read a post by Joe Clark this morning…

The W3C Closed Shop

Joe’s post — How Not To Fix HTML that the W3C are looking at fixing issues with HTML, despite the fact that HTML works pretty much okay at the moment, and when you consider that the vast majority of websites are designed either with no DOCTYPE at all, or possibly with HTML 4.01 Transitional at a push, it does seem odd to be focussing on this. On the other hand, WCAG 1.0 is widely used, and indeed case law is starting to develop to reflect the rights and needs of disabled users to access websites. It is widely acknowledged that the current guidelines are in serious need of updating, and yet no additional focus seems to be happening here.

Why is this? Is it because the W3C aren’t able to recognise the needs of the wider community? Is it because they’re not representative of the wider community? Is it because they barely acknowledge the wider community? Yes, yes and yes.

As Joe puts it:

…only W3C employees, Members (capital M sic), and invited experts get to vote. “Other stakeholders” can be and safely are ignored. Even invited experts can be and safely are ignored.Joe Clark

So, if you belong to a company that can pay to be a Member of the W3C, you can influence the standards of the entire internet. If you don’t belong to a company that can afford to pay — or if you’re a private individual — then at best you’ve got a right to express an opinion but your opinion carries no weight. Well, bollocks to that. If the W3C aren’t prepared to give web developers and web users a say in how they plan the web should be developed, then I don’t think the W3C have any right at all to tell us what to do. I certainly don’t remember voting for ‘em.

What we want is a democracy, not a dictatorship. We’ve even already had a bloody good suggestion of how to start going about it — Eric Meyer suggested the W3C become financially independent, which would help reduce the perception that the W3C is ruled by those companies big enough to buy in to it. Once we’ve done that, we can set up some method whereby developers and web users have some form of significant voting representation on their boards (the software developers and browser makers still need to be in there, but currently it’s too one-sided).

If the W3C truly want to be the standards body on the internet, they need to learn to stand on their own too feet, and not in someone else’s pocket. If organisation is financially dependent on other businesses then why shouldn’t we assume that their standards are not independent at all, but are directly influenced by those businesses? That is precisely what we should assume — because no-one other than those businesses gets a vote.

One way of demonstrating this is to look at WCAG 2.0:

WCAG 2.0 — Little Or No Progress

One of the things with WCAG 2.0 is that it dropped a lot of criteria that were in WCAG 1.0 because they weren’t easily testable. They may have been very advantageous to people with disabilities, but it was difficult to test for them, so let’s drop them. Why? I had thought that the primary purpose of WCAG was to help users with disabilities use the internet? Or is being able to demonstrate conformance to WCAG now of more importance than disabled people actually being able to use the internet? It would appear so. Which suggests to me that the focus is not where it should be — on removing barriers for people with disabilities.

I’ll just emphasise that I’ve got no gripes with any individual members of the working group; the only one I know is a top bloke, and I know he cares passionately about making content accessible to people with disabilities. I am objecting to the overall structure of the W3C, and probably some of the working groups decisions too — although I’ll hope you notice I do give them credit occassionally too!

The final date for comments for WCAG 2.0 was extended to the 22nd of June 2006. That’s now more than four months ago. They’ve received a lot of comments (even if you exclude the ones exhorting you to “download free ringtones”) and so it’s reasonable that this was going to take some time, but after four months you ought to be able to at least make some pronouncement on how long it’s likely to take.

So what have they been up to then? Well, I asked what’s happening in WCAG 2.0? on September the 5th, and answered my own question, so I’m just really going to look at what’s happened since. I’ll be referring mostly to WCAG WG minutes and to the basic issue search page, which allows you to search by issue number and actually attempt to make some sense out of the minutes, because as I’ve said before, despite WCAG 2.0 requiring definitions to be provided, the minutes just show things like “LC-844″ and expect you to make sense of it, when all they would need to do would be to link to their own issue tracking database with that ID number. I mean, it’s not like it’s hard to do, so why not do it? It’s either because they want the minutes to be as impenetrable as possible to discourage people from reading them of because they are too bloody lazy to do it.

Well, here and now, I’m prepared to make the offer that if they send me the working group minutes, I’ll put the hyperlinks in for them, and then send them the document back. It’s not hard. All I have to do is something like this: LC-844 and just change the parameter on the end each time. It’s really not that difficult. If they don’t want to do that, and still aren’t prepared to do it themselves, then I’m quite within my rights to presume that it’s because they deliberately want to discourage people from reading the minutes.

Well, I’ve read the minutes anyway and I even think I understand what they mean by them in some places. I’ll do my best to pick out a couple of points from each of them for ya…

WCAG WG Minutes 07 September 2006

Success criterion 2.4.4 has been clarified to mean that he purpose of each link can be determined from the link text (including supplemental and alternative text) and the text of any immediate context, which is defined to be the enclosing paragraph, list item or table cell (with any headers). That’s clearer than the original, so well done to all concerned for a sensible definition — and I’m sure Joe Clark will be happy to see that the context is taken into account!

The WCAG WG have decided to introduce an advisory technique relating to labels, to advise that the labels should be positioned in such a way as to “maximise the predictability of relationships”. However, unfortunately — unless I’m missing something, which is certainly possible! —there doesn’t appear to be anything in WCAG 2.0 that requires labels to be explicitly associated with their required fields. Surely therefore this is a step back?

WCAG WG Minutes 14 September 2006

The WG quite sensibly refused to be drawn on whether or not “acronyms” are a subset of “initialisms” as the two terms can be quite contentious — for example, most people would say that the term XHTML is at least one of the two, but since the first word of the term is “extensible”, it’s certainly not an initialism, and some people may consider it an abbreviation; others would term it an acronym by dint of common usage. The WG have sensibly opted not to get involved in this and pointed out that as both need to have an expanded version provided by some means, does it really matter?

WCAG WG Minutes 21 September 2006

The Working Group note have considered the “Hasta la vista”, baby” question — that is exactly when does a single word or short phrase need to be marked up the appropriate language.

Single words from one language are so often used in text written in another language that it is not possible to determine in a testable way when they have become part of the language or are still considered foreign words. Also, technical words and brand names are not translated and do not need to be marked.

—apparently short phrases will always need to be marked up otherwise they are still in breach of this checkpoint. C’est la vie!

WCAG WG Minutes 28 September 2006

It was suggested that if someone is automatically timed out of an authenticated session, and is not able to continue after re-establishing their security credentials, then this should be a higher priority than level 3 because it is likely to mean that the particular user in question simply cannot use that system. The Working Group explained that although they agree with this logic to some extent, they believe there are valid security reasons why this cannot be made to apply to all websites, and therefore this criteria belongs at level 3. I agree with the Working Group’s logic to some extent although would suggest that arbitrarily declaring that any criterion which does not apply to all sites belongs in Level 3 is the real problem here — why can’t you just comply with the critieria which are relevant to you? WCAG 1.0 was quite happy to use phrases like “…if you use tables…”. Why can’t we say here “unless there are legal reasons why…”?

WCAG WG Minutes 05 October 2006

According to the minutes, no items were discussed directly although in the “Editors Survey” item, there was a point which managed to say both “RESOLUTION: accept 749 as proposed” and “discussed with no resolution”. Anyone mind telling us which one?. Not that it really matters — it’s just another one of the endless discussions over the differences between a “web unit”, a “web page” and some “web content”. Excuse me while I get a “life”…

WCAG WG Minutes 12 October 2006

Again, no items discussed directly, although there was a significant clarification — they are going to provide a definition of a “web site” in the glossary.

WCAG WG Minutes 19 October 2006

There was an extremely useful comment about disabled form elements which are greyed out. It may appear as though the greying-out of the elements is a purely visual and colour dependent technique, but the Working Group have clarified that this is not the case:

Form elements which are disabled via markup or scripting, are greyed out and made inactive by the user agent. When in the disabled state these elements do not receive focus. Assistive technologies can programmatically determine the state of disabled elements and will provide this information to the user as the elements are encountered on the page. The change in color and loss of focus provides redundant, visual information about the state of the control.Working Group

The Working Group have also acknowledged a comment that the title attribute is not particularly well supported by assistive technology, but feel that this can still be used in advisory techniques — and they would hope to see improvements in user agent and assistive technology support of this attribute in the future.

The Working Group have also looked at the issue of text sizing and are suggesting something along the lines of:

Text size is specified using relative units that permit user control.

… and still nothing on validity, despite the fact that the Working Group received 18 comments asking for validity to be re-included as either a level 1 or level 2 success criterion. I would hope that the working group is actually going to listen to the web development community that it is claiming to represent. But that’s the problem with Standards By Dictat isn’t it? We’re all asked to contribute, but the contributions that the working group don’t like can simply be rejected or ignored. And that’s why we need individuals, disability groups, developers, and users represented at the highest levels of the W3C, not just those who represent big business.