WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Access Keys


From: John Foliot
Date: May 21, 2010 7:12PM

>> I have no use for them either and just wondered where such a
>> recommendation existed because of all the negative things I have read.
> If I am not mistaking it originated from UK Government. Then followed by
> Canadian Government however after the wats.ca articles then Canadian
> government as Jukka mentioned removed recommendation for using access
> keys.

Actually Davoud, the Canadian Government originally adopted Accesskeys
prior to the UK government as they were first recommended in WCAG1, and it
was our research at WATS.ca (we were doing a fair bit of consulting work
to the Canadian government at the time) which ultimately convinced them to
abandon them due to conflicts with Adaptive Technology. I also spent a
fair bit of time in the early 2000's writing and speaking about the issues
around accesskeys, which saw traction and thus they were not taken up that

Accesskeys *could* be useful were it not for implementation issues, both
w.r.t. conflicts with AT, but also issues of internationalization and
discoverability. My good friend (and a good friend of accessibility going
back over a decade) Chaals (a.k.a. Charles McCathieNevile - CSO at Opera)
has a proposal which needs finalization for HTML5, which would resurrect
Acccesskeys by insisting on a common browser means to query if a page has
accesskeys, as well as offering a means to re-map or otherwise interact
with those 'shortcuts', as this is what they are - shortcuts to inter-page
navigation. At the same time however, reliance on these types of
mechanisms will likely also have 'competition' from the new HTML5 landmark
elements, which will allow for users to also navigate from identified
region to region (certainly this will be significantly easier in
screen-reading technology, but might require some additional configuration
in the user-agents [browsers] to extend to other users - perhaps something
like a Firefox plug-in).

All things being equal in 2010, I still recommend authors avoid using