'

[Irl-dean] WCAG 2

Joshue O Connor joshue.oconnor at ncbi.ie
Thu Jun 1 10:53:52 IST 2006


Hi Tim/Barry,

> Um ... to Tim (and others) I would have to recommend against
> taking Joe Clark "at his word".  He certainly does have axes to
> grind.  I would say that there is some good sense and validity in
> what he says about WCAG 2.0, but it should certainly not be
> accepted uncritically either.


Thats a fact.

> Yes this is true (in a sense); but it is *not* new with WCAG 2.0.
> WCAG 1.0 always allowed a site operator to specify the "scope" of
> any conformance claim. 


Yes, as users could only claim conformance for a particular page and the presence
of a WCAG badge on a page or "web unit" certainly does not mean conformance for an entire website.

> The significance - and acceptability - of that, then depends on
> the specific reasons for the site operator being interested in
> WCAG conformance at all. 


Well thats an issue of motivation. Is the author merely jumping through the hoops
so they "appear" to be accessible or are they truly trying to to the best they can to create accessible systems? 
How do you quantify that? If you cannot quantify it,
then to my mind it should be mandatory that *ALL* pages, web units or widgets,
are accessible (where possible), full stop. Then it doesn't matter about the site operators *reasons* because in the final analysis they will have to provide an accessible service.

> To take a different case: imagine a commercial talk radio station
> that decides to make selected programmes available as podcasts
> via their web site. 


Thats a good example, and one that is timely. Under the new WCAG are podcasts considered multimedia anymore?
They are an audio format that will serve practically everybody, except those who are deaf
or hearing impaired. As they are not a combination of media, no pictures just sound, are they
considered to be multimedia and should they then be treated accordingly by the new WCAG 2 guidelines?
with as Barry mentions, transcripts or text versions being made available.

> So what should the station do?  One option would be to simply not
> provide the podcasts at all!


That would be terrible situation, if due to an ill-conceived guideline that is impractical  or
difficult to implement , the line of least resistance is just not to use podcasts?   I can't see it happening, as the world doesn't operate like that. As William Gibson said, "The street finds its own uses for things", but the possibility that it could is alarming and stultifying.

IMO, It would be better to, using this example, continue to podcast and if some users require podcasts
or transcripts in a different format then they can be supplied as needed. This is practical. Another similar
example is PDF documents. If an agency has 100,000 legacy documents in PDF format, they may not
have the resources to convert the bulk of these documents into accessible "tagged" documents.
So what is the most practical solution? Convert the  documents that are used the most into accessible tagged
formats, and when requests for archived documents are made, then convert them as  needed.

> So, in my opinion, "non-discrimination" should not mean that we
> cannot deliver a service to "anybody" unless we deliver it
> ("equally"?) to "everybody", regardless of what burden that may
> involve. 


I'll buy that.

> I also sincerely believe that it does a
> disservice to the equality cause to pretend that we live in a
> world without such dilemmas and judgement calls.


This is the nub of the problem with the successful application of so called "standards". 
There are often many shades of grey in real world situations that need measured judgment calls and compromises. For me, I am more interested in "practical" accessibility. Many people with disabilities are on limited budgets and using older equipment that just won't fly with "development".

> *WCAG* can't be expected to deal with every possible
> such situation.  WCAG is just guidance on what would be "best
> practice": no more and no less. 


Thats where I am concerned that it is "loosing touch". For example, (if I have this right) the idea of comparing User Agent DOM outputs puts the  responsibility in the hands of the authors to write code that works with the *output* of a particular user agent, and seems to indicate a move away from the (good practice) of creating valid code. The concern here that we are (again) moving towards the
kind of "quirks mode" development that was a nightmare in the 90's.

<q> objective of this technique is to prevent Web pages from being presented inconsistently 
to users whose assistive technologies utilize a document object model (DOM). 
Incorrectly nested markup leads to inconsistent DOM representations being generated by user agents. 
The HTML specification does not define how user agents should deal with incorrectly nested markup, so user agent manufacturers have to define their own methods of dealing with it. In an attempt to honour the author's markup, some user agents introduce extra elements in the DOM to maintain a tree-like structure; other user agents end up with a DOM structure that is no longer a tree. 
As assistive technology gets more sophisticated and starts to make more use of the DOM, consistency becomes more critical.</q>

Dealing with DOM inconsistency should be the realm of the User Agent manufactures them selves ( I could be wrong but the implication
here is that it will now be the responsibility of the content author). 

Please correct me if I am wrong on this. See

http://www.w3.org/TR/WCAG20-TECHS/#F28-procedure

Now, this could a practical  move in the right direction that I just don't "get". Its possibly an attempt to compensate for the lack of user agent support for various coding languages etc.

As Barry says WCAG is made up of many of the the brightest, and after 5 years of chewing the fat with this they must have
good reasons for the suggestions they have made. I have my own concerns about the commercial, vendor driven nature of W3C,
but I don't think that Joes idea of the WCAG samurai is a good idea either. It's divisive, and as he is a voice that many in the developer community listen to and respect (and in many ways rightly so) then the onus is on him to keep his personal
grudges under wraps and try to deliver a fair and balanced message that is in itself accessible and serves the development
community in the spirit with which he supposedly champions.

Josh

Joshue O Connor

Senior Web Accessibility Consultant

Centre for Inclusive Technology (CFIT)
National Council for the Blind of Ireland

Website: http://www.cfit.ie
E-Mail: joshue.oconnor at cfit.ie
Tel: 01 8821980



Barry McMullin wrote:

>
> On Wed, 31 May 2006, Tim Culhane wrote:
>
>
>> Note firstly that I haven't  read the actual WCAG 2.0 documents,  since I
>> don't haven't to have a free couple of days at the moment.  Given this I am
>> taking Mr. Clarke at his word,  and I'm assuming what he says is true and
>> accurate, and not just the output of somebody who has a personal axe to
>> grind.
>
>
>
> Um ... to Tim (and others) I would have to recommend against
> taking Joe Clark "at his word".  He certainly does have axes to
> grind.  I would say that there is some good sense and validity in
> what he says about WCAG 2.0, but it should certainly not be
> accepted uncritically either.
>
> For example:
>
>
>> 6. You'll be able to define entire directories of your site as
>> off-limits to accessibility (including, in WCAG 2�s own
>> example, all your freestanding videos).
>
>
>
> Yes this is true (in a sense); but it is *not* new with WCAG 2.0.
> WCAG 1.0 always allowed a site operator to specify the "scope" of
> any conformance claim.  Further, given that WCAG, per se, is
> simply a voluntary guide, this is simply common sense.  If, for
> whatever reason, a site operator can't or won't make certain
> parts of a site conform with WCAG, they should not be
> *discouraged* from making all the *other* parts conform.  And
> that's as far as WCAG can reasonably go.
>
> But be clear that, where a site operator goes down that route,
> they can no longer claim that the "site" conforms to WCAG (1.0 or
> 2.0).  They can only claim that some, clearly identified part(s)
> conform.
>
> The significance - and acceptability - of that, then depends on
> the specific reasons for the site operator being interested in
> WCAG conformance at all.  But to take our local case, if the NDA
> Code of Practice on the Disability Act 2005 is actually endorsed,
> then that would, for most practical purposes, mean that Irish
> public sector websites are required to conform to WCAG 1.0
> Double-A. Period.  Not some parts of the sites, but the full
> sites.  But *that* requirement is - quite properly - not dictated
> by WCAG itself, but by the particular local situation,
> legislation, etc.
>
> To take a different case: imagine a commercial talk radio station
> that decides to make selected programmes available as podcasts
> via their web site.  They are not a public sector body so the
> Disability Act 2005 does not apply.  But the Equality Act (for
> example) does, and they must not discriminate on grounds of
> disability.  WCAG 1.0 conformance would require them to make (at
> least) transcripts of the podcasts available.  Note that this is
> a completely new requirement, that their "primary" business
> (broadcast radio) does not require.  But, generating transcripts
> could be a quite significant additional financial burden in that
> case.
>
> So what should the station do?  One option would be to simply not
> provide the podcasts at all! If they don't provide the audio,
> they have no obligation to provide transcripts, and could achieve
> WCAG conformance in the narrow, artificial, sense.  But I'm
> presuming that nobody would consider that a reasonable solution
> (?).  It would simply deprive a huge number of people - including
> many with a wide variety of disabilities - of a very useful
> service.
>
> Another option would be for them to, in good faith, consider how
> much (if any) transcripting they could "reasonably" afford to do;
> but to then explicitly exclude the remaining podcasts from their
> claim of WCAG conformance.
>
> Would this be legal?
>
> Quite possibly - the Equality Act requires "reasonable"
> accommodation - but no more than that.  In practice, if someone
> complained, the site operator would have to make a case for why
> their behaviour is "reasonable", but, if they can show that they
> didn't act arbitrarily - that they are aware of the limitation
> (which of course they must be, if they explicitly restrict the
> scope of their WCAG conformance claim) - and that they made a
> genuine, good faith, assessment of what they could "reasonably"
> afford to do, then, at the very least, they would have a
> "reasonable" defence to any complaint of discrimination.
>
> So, in my opinion, "non-discrimination" should not mean that we
> cannot deliver a service to "anybody" unless we deliver it
> ("equally"?) to "everybody", regardless of what burden that may
> involve. Perhaps not everyone here will agree with me on this;
> but, while I do have some worries about abuse of the "reasonable
> accommodation" test, I also sincerely believe that it does a
> disservice to the equality cause to pretend that we live in a
> world without such dilemmas and judgement calls.
>
> Anyway, the point here, of course, is not to say whether in fact
> the particular choice in this example would be legal (or, indeed,
> ethical, which is a somewhat different thing). It's simply to
> show that *WCAG* can't be expected to deal with every possible
> such situation.  WCAG is just guidance on what would be "best
> practice": no more and no less.  It cannot, in itself, compel
> anybody to do anything.  So it would make no sense whatever for
> WCAG to *insist* on an "all or nothing" kind of conformance.
>
> To take the example above again, surely we do *not* want to say
> to the radio station: "Look, your podcasts (without transcripts)
> mean your web site cannot conform to WCAG, therefore you might as
> well ignore WCAG completely!"  It would surely be much better to
> be able to say: "Well, you have made a good faith decision that
> you cannot afford to meet this particular WCAG guideline, for
> these particular resources.  I may (or may not!)  agree with your
> decision there.  BUt I would certainly *still* advocate that you
> make *everything else* on your site conform to WCAG!"
>
> OK, I've gone on at some length to tease out that example (and,
> lest anybody has missed my usual disclaimer, "I am not a lawyer
> and none of this is legal advice"!).  You may or may not agree
> with me, but that is not really the issue.  My point is simply
> that Clark (in his inimitable style) has taken something that is
> genuinely part of WCAG 2.0 (well, actually, part of WCAG 1.0),
> but presented it in a simplistic, out-of-context way, that makes
> it seem just silly.  Whereas, in fact, it is trying to address
> something which is complex, subtle, and hard to deal with.  One
> may or may not agree with everything in WCAG 2.0 (I also have
> doubts or concerns about some aspects of it); but it is unfair to
> imply that the WCAG working group are somehow completely
> mis-guided or out of touch.  They are not, and every single line
> of WCAG 2.0 *has* already been subject to long and critical
> study.  They may still have got it wrong, but certainly not
> "casually" so!
>
> Of course, the real sting in the tail of Clark's piece is his
> proposed "solution" - being to put together a self-selecting and
> secret band of "experts" to put together a "WCAG 1.1" instead.
> Now I am as impatient as the next person with the delays and
> cumbersome process that WCAG 2.0 has been subject to. I am also
> uncomfortable as to whether W3C is really an appropriate body to
> be doing this anyway (it is very hard for W3C to avoid at least
> the *appearance* that its commercial sponsors may have an
> inappropriate influence).  But, imperfect as it is, I certainly
> would not swap it for the kind of process Joe Clark proposes!
>
> Best - Barry.
>
> _______________________________________________
> Irl-dean mailing list
> Irl-dean at list.eeng.dcu.ie
> http://list.eeng.dcu.ie/mailman/listinfo/irl-dean
>
>







More information about the CEUD-ICT mailing list