'

[Irl-dean] Creating accessible javascript

Barry McMullin mcmullin at eeng.dcu.ie
Sat Sep 30 11:59:41 IST 2006


On Fri, 29 Sep 2006 drice at nda.ie wrote:

> The page in question "Use JavaScript wisely" [...]  starts with
> "Use JavaScript by all means...".  We never intended to give
> the impression that JavaScript is automatically 'bad' for
> accessibility.  The statement "Reliance on JavaScript is an
> accessibility issue in itself" does not reflect the nuances you
> outlined.  However it is an issue where the browser does not
> support it or it is turned off.  Still, good points - will
> review and amend.

Sorry to keep flogging this one, but it is quite subtle, so I'd
like one last (?) bite at it.

It seems to me that there are quite two distinct points that
might be being made here:

+ *Using* Javascript is not *necessarily* bad for accessibility,
  and may even positively enhance accessibility for at least some
  users;

+ But *reliance* on Javascript for any "important" content or
  functionality *is* necessarily bad for accessibility.

I think we are all agreed on the first of these; and we are even
all agreed that reliance on Javascript is a bad thing "in
general" (because users may be unable or unwilling to permit it,
but for reasons that need not have anything whatever to do with
disability).

So the only possible lingering disagreement is whether reliance on
javascript is a bad thing on any specific *accessibility* grounds
(i.e. related to disability).

My own view is that it *is* bad on specific accessibility grounds
- basically because, in the current state of the technology, it
is difficult, verging on the impossible, to guarantee that
javascript will not actually interfere with the behaviour of
(any) assistive technology.  Therefore, while people may disable
javascript for reasons unrelated to disability, they *may* also
disable it for reasons that are indeed directly related to
disability. And *that* makes it an accessibility issue,
regardless of any other considerations.

This view matches the NDA draft text which Dónal quotes above
(and I had absolutely nothing to do with writing it!).

Further, this position is currently exactly compatible with WCAG
1.0 - Checkpoint 6.3 - and, I would argue, with UAAG 1.0, which
requires only that user agents offer certain APIs, intended to
support client-side AT, but does not at all require user agents
to actually execute any scripts at all, never mind arbitrary
scripts provided from the server side. Compatibility with WCAG
2.0 will presumably depend on one's attitude to, or choice of,
"baseline".  But I will argue, precisely on the grounds indicated
above, that javascript should not be included in any WCAG 2.0
baseline at this current time.  If one accepts that position,
then the advice under WCAG 2.0 will still be exactly the same -
one should not *rely* on javascript.

Of course, to re-iterate one last time, that absolutely does not
mean that I am recommending against the *use* of javascript. On
the contrary, it remains a very useful and important technology with
significant potential accessibility benefits, and certainly
*should* be used in any appropriate circumstances.  I am arguing
only against allowing *reliance* on javascript...

Best - Barry.




More information about the CEUD-ICT mailing list