Thursday, 29 January 2009

What to do With Accessibility?

Anyone who hasn't been on Mars will know that the W3C has just released version 2.0 of their Web Content Accessibility Guidelines. In due course these will become a British Standard (BS 8878) (draft here - which apparently goes past it's sell by date on the 31st...).

One simple thing that's worth stating is that the speed of the web and the speed of any committee are so far apart we can all be pretty confident that the new guidelines will be out of date in some capacity as soon as they were confirmed. W3C have had a fair amount of trouble trying to strike the balance between practical and content-agnostic, I'll leave it to you to decide how well they did.

If you want a solid overview, there' a great video introduction by Paul Boag here. My suggestion is that you go watch that, or if you're brave/underworked go read the full spec here. It's ok, I'll wait...

Great, - who knew that WCAG is pronounced 'whuk-ag' eh! 

There's a few interesting points which have real relevance to anyone wanting to provide Web 2.0-y services - and a few questions that popped up in my mind.

Principles vs Practice

Here's an issue I have with accessibility guidelines: they start off with a straightforward and fair principle (such as "Everyone should have easy access to web resources") and then try and draw up rules or approaches which instruct you in how to meet that principle. The problem is that, in reality, the idea of 'universal access' is so wide that it's impossible to pin down - running the gamut from sensory disabilities (blindness, deafness) through to varying levels of literacy problems. 

At the end of this attempt to tie down such a wide topic to a list of 'dos' and 'do nots' you're then forced to go back and correct the principle in order to make it fit the inherent flaws in the practice (WCAG2.0's is  to"...make Web content more accessible to people with disabilities" - note the more). Wouldn't it just be better to trust organisations to meet the original principle, and to explain how they do (and are striving to) individually?

Ok, ok, if we let everyone choose how they want to meet the principle of "accessibility" then there will probably be a huge proportion of people who just ignore it... oh wait, doesn't that happen now?

Content vs Technology

One of the best aspects of WCAG2.0 in comparison to version 1 is that it's much more content focussed. Version 1 was, lets face it, written with the idea that the W3C is the guardian of the Internet - only W3C technologies are supposed to be used and the assumption is that they would be able to to keep up with the Web (ah, didn't we all think that last Millennium?).

So, WCAG2.0 accepts that - yes - the technologies will change and allows you to take advantage of that whilst still retaining some level of accessibility (by allowing you to rate content individually rather than as a single blob of web content). So the language and terms you use are as (if not more) important than the mere technology which provides them. 

What this means, in practice, is that everyone who writes for the Web needs to be thinking about accessibility issues from the second they put fingers to keyboard (no, this isn't the case at the moment for most organisations - think about all those people who write in Word and email it to someone to copy/paste into your CMS...). Do your content writers have guidelines?

People vs Programs

One aspect of WCAG(1 & 2), which is iterated more clearly in BS 8878, is the idea that real accessibility is best achieved by testing sites with users who have varying levels of disability. In my view, there's no other option - sure I can use a screen reader and fire up a contrast testing program but I'm not really judging the entire process from my own experience - as a disabled person would.

I'd go further to say that rather than seeking out specific groups who offer 'testing services' we should really be seeking out people who use our sites - what better evaluation can you get than from your own users who will give feedback on the context of the feature or content you're testing and not just the specific 'tick list' that most consultants provide.

Controled vs Cloud

Here's a question which every organisation is going to face in the next few years - where do our responsibilities in regard to accessibility end? It's no longer the case that we keep to our own web domains, store our servers two floors down and have total control over every piece of our web estate.

Even if we take every step to keep our hosted sites AAA compliant what level of responsibility do we have over our spaces on Flickr or Facebook? For these sites it's the market that drives their need to cater for various access needs - and we cannot guarantee that they will always cater for the kinds of requirements WCAG or British Standards require.

Cost vs Benefit

Now here is a tough one. What kind of steps should we take to make our web content accessible? How do we, pragmatically, ensure that we fulfil our legal and moral obligations whilst, at the same time, not falling foul of 'designing for the minority'.

How simple (technologically and linguistically) should we make the web? More importantly, if we choose not to sacrifice some aspects of our approach how do we explain this to users?

BS 8878 (in draft) says:

"Any user claiming conformance to this British Standard is expected to be able to justify any course of action that deviates from its recommendations."
but what form of justification isn't made clear. WCAG suggests that we:
• Consider using a combination of technologies to achieve your goals.
• If a technology meets some but not all of your criteria, consider providing an alternative solution, which provides disabled people with access to the same content, updated at the same time as the original content.
• If a technology meets none or only a few of your criteria, and an alternative, accessible method of providing that content is not reasonable or possible, ensure you let your disabled users know why this is the case.
In many cases we might seek to make content available to a wider audience by putting it on the web. Take video or streaming of talks and training sessions for example.

An event which might have been 'accessible' only to 20 or 30 participants who can physically make it to a room can be streamed or, better yet, recorded and posted online for very little cost. At this point the content is obviously more accessible than it was because it can be accessed by more people.

On top of that, the content is no less accessible than it was when the physical event took place (it's still lacking audio description or textual transcription). Adding transcriptions or other accessible assitive content can often cost far more than the original event.

The alternative for many is just to not put such content online - is that really the best solution? Is universal equality through universal poverty of access really acceptable?

Perhaps another alternative exists, Paul Boag's video has a community created transcript. If we provide the means for crowdsourced additional information perhaps we can, then, bridge the gap between wholly inaccessible (ie offline) and partially-accessible (non-WCAG compliant) content?

A Final Note

Often the reaction to the potential legal ramifications of the Disability Dissemination Act (and thus WCAG and BS guidelines) is disproportionate concern. Meeting the key principles of access, and justifying your approach, will probably negate the worries of most users - and the most important thing (far more important than meeting the specifics of W3C and BSI documents) is that we listen to users when they talk to us about our websites.

Accessibility guidelines can also tend to be manipulated, or oversold, by companies eagre to sell their accessibility services and by web designers who would rather not have the hassle of learning a new technology or improving their own skills.

PS: Brian Kelly (all round UKOLN web guru) has some great posts on accessibility over a fairly long period - his ideas about Accessibility 3.0 are interesting, he seems to be erring more toward the principles approach - and the application of API-type approaches to solve access issues. His other posts (1234) are all very interesting too - particularly the last post on Adaptability.

No comments: