WCAG2 - A little light reading

I was getting to the end of this long but very well researched, well thought-out and well written article, when I thought: maybe we ought to get together and publish our own 'real world' accessibility guide and give it a fancy title and logo. Then I read this...

"Now, though, I can announce that such errata really are going to be published, and my friends and I are going to do the publishing. After the manner of Zeldman’s CSS Samurai posse, which put CSS layouts on the map for browser makers and developers, the WCAG Samurai will publish errata for, and extensions to, existing accessibility specifications."

That sounds good to me. I'll wager that the elite 'standardised' community to which many of us aspire to be members, will happily ignore WCAG2 in favour of something more meaningful. As for the governments and corporates that specify WCAG2 conformance: perhaps a link to the above article might help!

See comment number 22: The problem runs deeper

http://www.alistapart.com/comments/tohellwithwcag2/?page=3

...an interesting observation!

All rather too depressing, the insightful comment #22 tends to re-enforce a general gloomy feeling I've felt for sometime in respect of these matters - not the general concepts- but the direction and motivation of people/corps controlling things, but I'm working backwards reading comments first , too late to read the links proper I shall save that for the morning along with methinks a necessary 2 or 3 cups of strong coffee.

Hugo.

Roger Johansson has written about it as well with a few links to other articles on the topic. (He's not been invited to be a part of the Samurai.)

what a depressing state of affairs. when you're battling all day at work with those high up to get them to take accessibility seriously the last thing you need is this kind of madness....

:sigh:

I did indeed sit down this morning and proceed to read through all 81 comments to this article , and naturally the article itself.

To a large degree I feel that there is probably a lot of truth in some of the negative points being directed against the wcag and it's working group and the whole supposed ethos that the W3C rumbles along under, and indeed I did spend time jumping up and down getting hot and bothered over this state of affairs , if as purported we are about to be plunged into a new age of unreason, as that would be an utterly disastrous state of affairs.

However reading the comments tends to remind one not to be hasty and to look for balance and sanity amongst this furore.

Reading Joe Clarks main bullet points of contention again, I worried that they seemed to be somewhat bombastic (for want of a better phrase) so I wondered whether in actual fact I was reading an accurate summary of these issues, and I have to admit that I think not and that it's important to possibly try and interpret the guidelines(Success Criteria?).

Joe Clarke asserts that

(paraphrased)

Reading the section nowhere can I find mention that future sites won't require valid HTML, instead I find a fairly sound piece of thinking that seems to say that " DOM structure output should be compared in two or more UA using DOM inspectors in order to compare differences, this due to the fact that HTML does not specify how browsers should deal with illegal markup, using a DOM inspector will allow you to see how the browser has adjusted the Document Object Model to reflect the changes it deemed necessary to render the page.

Therefore the DOM is the best method of actually determining browsers are actually 'seeing' the same code and constructing the same Object model.

Assistive devises start to make more use of the DOM and therefore it becomes a useful tool in ensuring that these devices see the same thing, not that 'It's ok to have invalid HTML'



Was there ever an actual absolute prohibition to using tables for layouts?
Regardless; here they do not appear to be saying you can use tables for layouts but are using a table and its related elements as an example of how it would be incorrect to use it.

The actual section describes the error in misusing elements such as h# tags where one wants large bolded text and where the content does not describe a new section of the page and where a heading tag should be correctly used, it goes on to clarify that it does not include using a table in this example as an error UNLESS one were to use it's related elements such as TH or Caption elements which would mean that you were attempting to present the content, to describe a relationship that didn't exist in this context.

What actually is being said here is that tables for layout does NOT exemplify the error being described in this section NOT that one CAN use tables for layouts.



Whilst I do perceive an intention on the part of large companies with new technology to allow themselves some form of dispensation in the use of their technologies, actually this sounds a sensible provision allowing the developer to state that certain parts of a site may not fulfil certain requirements, it does not say that users have no recource to complain about that level or that the developer can make a blanket statement about their site and be safe from the requirements of a countries legislation.

These are examples of where possibly Mr Clarke has been perhaps a little eager to place his interpretations on things with perhaps just a little bit of clouded judgement, then again he is not without some experience in these matters and a lot of his vexations towards the working group do seem to be borne out by others with perhaps equal if not greater first hand experience of this closed world.

Hugo.

My 'interpretation' of his article was that was what he saw as part of the problem: you've drawn one set of conclusions about WCAG 2.0 and he's drawn another. A third person might come up with something different again.
I got the impression that he (and others) were hoping for a bit more clarity and a bit more hard and fast standardisation than is being offered.

Three people ... three opinions.

What struck me most was the mechanisms that worked against those the standard is aimed at participating in its development and that the final product is not likely to advance web accessability - at least not by as much as was hoped.

Hugo highlighted some points that are close to our hearts as web developers, but the main issue has to be using the available technology to make the web accessible to all those who struggle with the web when its content is delivered using the #1 paradigm - monitor/mouse/windowing web browser. Joe seems to be saying the standard has copped out in providing that. Other comments broadly agree with him and the one defender of WCAG2, for whatever reason, picked specific points to argue on without really defending the worth of the final product.

It'd be good to hear from Lorraine as whether she believes WCAG2 will advance web accessibility and if it does will it be by as much as it could/should.

In truth I draw the same conclusions that the detractors do on this subject and to a very large extent was entirely unsurprised by what has been delivered, I have long thought that the W3C is failing on many levels and most definitely in the area of clarity, and I have moaned in the past about the inability of the W3C to write plain clear english.

My previous post was less about my conclusion and more to do with the fact that Mr Clark was stating things in a particularly polemic way.
There will always be a number of people that will read such an article and take it at face value, if the problem is that the WCAG 2.0 is open to interpretation or is less than specific, then it seems Mr Clark is making - in his bullet points - some very definite conclusions, that can be dangerous in it's own right.

From a brief perusal of 2.0 I agree totaly that it is a document that does nothing to further standards and accessibility, and just leaves me feeling ever so slightly confused and reluctant to espouse accessibility tenents; I found confusion in the vague ramblings about the nature of document flow/structure and the use of CSS to apply control to the source order, it left me actually doubting what I was working on and questioning how I was constructing a layout which was already a pain.

This sort of confusion I can well do without and find it totaly unacceptable , I have not got time for these sorts of problems I have to know what I'm doing and why, I'm not spending time I haven't got attempting to interpret requirements.

So yes I'm sure Mr Clark was hoping for a bit more clarity and standardization, I think we all were - those that were aware of the failings of 1.0 - and lo we seem to have taken a massive step backwards, great, 5 years to formulate a clear direction and set of directives and we get something vague and open to interpretation; that is just not good enough!

This would also seem to have implications in law if the WCAG is to be used as the yardstick and general guidance document as it is supposed to be under the EU legislation.

Ultimately this debacle will have the effect of driving a wedge into the standards, I agree that 'Samurai' is necessary but it's far from a satisfactory state of affairs.

All of this just tends to highlight the truth that no one seems keen on talking about that of the W3C tending not to deliver in many areas, CSS specs being one such example, you simply can not take so long debating and discussing things only to come up with specs that fall short of what people require "There is no provision for vertical centering in CSS2 but we will probably do something about that in CSS3" why was it not considered earlier ? who do these people serve ? is it in the draft for CSS3?

Law makers can be no better than standards writers. It will depend how much they also consult with the people for whom web accessibility is important.

I think its complex. Standards can't be imposed - at least not easily. A bad standard will simply be ignored. And standard development seems to suffer from chicken or egg. Until there is a chaos of different implementations the need for a standard is low. Building a standard before the chaos may mean you end up with a standard for something nobody needs.

Web browser development has slowed since the heady days of Navigator v IE. So it seems has standards development. Roll on another browser war!

http://www.w3.org/TR/2005/WD-css3-layout-20051215/#vertical

For some insight into the process, read the top portion of the CSS2.1 Working Draft, in particular, Candidate Recommendation Exit Criteria and Features at Risk. Unless two publicly available UAs (normally browsers) correctly implement a feature it'll be dropped.

Without clear guidance everyone will be floundering, law makers will consult those that accessibility is important to , they will know what they want but at the end of the day law as to prove one way or another that there has been a failure to comply with a set of guidelines , those need to be clearly defined or every one is just going to argue endlessly, the only ones to benefit will be the lawyers and their disgraceful charges they impose.

I think you can probably apply a little standards to something even before the apparent ned for order from chaos, but agree in the main standards arise from a need; "roll on another browser war" indeedy
I realise that shrangli la does not exist, things will never get any better than this, we shall always somewhat be dealing with a certain amount of chaos.

I was being somewhat rhetorical in those css comments not necessarily looking for insight I have actually read a good deal of these documents, but it does highlight some of the nonsense so MS won't implement certain properties as they consider them to be poorly specified and the 'Candidate recommendations' wait for UA to implement properties before considering them, yes I'm sure I am taking liberties, but those links sum it up for me.

No and... err... no.
I have given up slavishly following "standards" whatever they may be, howsoever interpreted by whomsoever. I am now plowing my own furrow, hence, long periods of absence from the forum. Whilst not exactly or totally ignoring standards I'm using "techniques" that, in my opinion, make for a successful site and are relevant to accessibility issues.

Eerily, even as a woman driver, I stay within the speed limit "standard" but... if I am driving downhill within the speed limit and an out-of-control juggernaut, designed and driven recklessly by a committee that forgot to include a brake pedal in the blueprint, is fast catching me up, do you think I would stay within the "standard" speed limit? Would anyone? :whistle:

I hope you can drop by from time to time to share your techniques with us.

I would like to "plow my own furrow" in a pragmatic sense. The trouble is, I don't have a broad knowledge of just what the issues are. I have a certain degree of confidence in Joe Clark from reading his book, Building Accessible Websites. His 'do the right thing' approach appeals to my sense of ethical pragmatism. Lorraine appears to have the same, or close to it, outlook. (If I missed your point, L, please correct me.)

I find the (x)html and css specs to be very clear and specific, having very few ambiguities. That makes coding a pretty straight forward task except for deciding what something is, semantics, or how something is constructed, er structure—is it a ul, dl, or table, eg.. UA shortcomings are the primary coding issues, and they are mostly clear cut. These things don't particularly, or directly affect the visitor.

Accessibility issues are about an individual's shortcomings. The fixes are not clear cut, to the developer, in the manner of IE's. An accessibility standard should, by my reckoning, take a two pronged approach.

1. First, it should address each class of disabilities and point out the things
   1. that generally affect that disability,
   2. the things that are harmful or throw up barriers,
   3. and finally, the positive steps that may be taken to enhance access.
The web developer will have a lucid set of issues from which to work.


2. Second, there should be a three tiered validation test. Many of the test queries will be objective, "is alternate text provided on all images", and many will be subjective, "is the alternate text worth a hill of beans". To "pass" a level's test, the objective issues should be checked as satisfactory, and the subjective? That will depend on a good faith effort. :shrug:
   Very much like WCAG1, the tiers would be something like so:

1. A minimal allowed level. The checklist would test that there are no gross barriers to access. There is no requirement for accessibility enhancements. There's nothing to block access, but it's not necessarily easy for some arbitrary class of handicap.
   Pass the objective part with 100% and the subjective in good faith, and that's level 1 accessibility.
   
   
2. A generally fully accessible document. Not only are there no significant barriers, there are positive steps taken to enhance the handicapped visitors' experience.
   Pass levels 1 and 2 with 100% on the objective and good faith on the subjective, and you're level 2 accessible.
   
   
3. Finally, a level for the terminally analistic Utopianist. This is the wish list of items that individually seem desirable, but in practice contradict or break another item, or are just plain invalid.
   To gain level 3 compliance, pass levels 1 and 2 with 100% on the objective and good faith on the subjective, and at this level, the passing grade might be some degree less than 100% on the objective with good faith on the related subjective issues.

You'd think a large group of smart people, in five years, could pound something out. If there are problems that arise, well that's what bug fixes and minor revisions are about.

The usual W3C standards are aimed at the manufacturers. We (coders) read them to understand how the UA is supposed to respond to our markup. The accessibility standards apply to the web developer. We require a more transparent standard, with plain language explanations that are a part of those standards. At least, I do.

A prof told me (referring to one of my essays) there are two reasons for poor writing; you don't know the language, or you don't know the subject. I will add a third; you're purposely obfuscating or blowing fuliginous matter. I don't know the reason in this case, but even for a legalistic tome, that draft is especially obtuse and seems particularly counter to the base purpose of the recommendation.

for what it's worth,

gary

[/][/][/]

Gary - very well written Sir =D>

Lorraine - please share your thoughts with us some time regarding 'Lorraine's Practical Accessibily Standards'

WCAG2 is getting far too much cybertime at the moment... it just does not deserve it. People who should be otherwise gainfully employed in producing good looking, relatively barrier-free websites are agonising pointlessly over what appears to be the poorly collated minutes of five years of acrimonious committee meetings.

The argument for coding and accessibility standards is well founded. Many developers try to adhere to them or at least those they understand. But what of the mega-sites that receive millions of hits a day? Where are their standardistas? Certainly not in the board-room or finance office. Is it a co-incidence that so many of those sites are said to be in-accessible? "One day", we say, "they will find themselves in court, you just mark my words young fellow my lad!". I don't think so.

In the UK we have the Disability Discrimination Act and now PAS-78 to protect "vulnerable" surfers. But I seriously doubt we have organisations and individuals (with sufficient intestinal fortitude) that can cut through the current accessibility standards cruft and make a cohesive case to put before a learned judge who, by the way, still needs a dedicated clerk to operate his laptop in court. :?

Gary makes the most relevant distinction... coding standards are for machines, accessibility standards are for people and you need to understand how people interact with their machines to go down that route. So then we get into the realms of usability for mainstream surfers... another can of worms and ever so slightly OT.

Accessibility for all manner of impairments can seem like an insurmountable task and one that is almost impossible to standardise and assess given that so much of it is subjective. But, one step at a time, it is possible to produce sites that cater for more and more people. Years ago I found this article quite useful. It is rather old now, very basic and talks about how individuals with differing impairments interact with computers. It may be useful for developers starting out with accessibility, but comes with a health warning for standardistas... you are advised to be seated before reading and have a strong drink close at hand.

I am very lucky knowing many people with various levels of blindness who are more than happy to test sites for me (subject to a modest consideration, of course, if they test a commercial site). Three or four years ago I made contact with organisations for people with restricted mobility and also with various senior citizen's clubs so that I could better understand their computing problems. Now I am working with people who for want of a better term are "word-blind". In my opinion, these folks are the "also-rans" in the accessibility sweepstake. I'm having a hard, but enjoyable, time getting to grips with the problems experienced by people who are dyslexic, have cognitive/learning difficulties, read by symbols, need audio transcription etc. And I've made a lot of new friends.

It's important to get to know your audience and if you can also help out at a local club( for people with disabilities so much the better all around. You won't regret the human interaction, it will get you away from the keyboard for a while and it may even make you a better developer.

Lorraine the Pragmatist
(sans-standards, ops populus... err I think, perhaps our resident Latin scholar will correct :mrgreen: ).

Test your level of accessibility knowledge. Classic Joe Clark.

I consider myself HTML and CSS level 5, and between 4 and 5 on accessibility.

THis bit made me laugh though:

Is aware that spacer GIFs are tedious for screen-reader users to listen to and wouldn’t be caught dead subjecting a blind person to something as jejune as a filename. So alt="spacer!" it is. Aware of the controversy involving table usage; developer’s layout tables are fully qualified with summary="This table is used for layout purposes only". (Such declaration is duplicated in a table title attribute that helpfully follows the mouse cursor around the page.) Uses XHTML because it’s “better,” but hasn’t figured out how to save a Unicode file on their Windows system yet, turning a simple copyright symbol into a string of comic-book profanities. Is a cochair of the WCAG Working Group.

what I don't understand is that in 4 days we've had an interesting and insightful discussion with points that make sense. Maybe I'm naive but I don't understand why this is not possible at the W3C.

Maybe Roy's right and we should have our own accessibility samurai... we might actually get something usable from it!!!