Ever stop to think, and forget to start again?
Accesskeys: Where Are We Now?
Access keys are a way of defining keyboard shortcuts which, in theory, allow users to jump to, for example, the Search item on the web site navigation menu by selecting ALT+s. Depending on the browser being used, the user may then have to press ENTER to activate the link. The designer can define which keys, in conjunction with ALT (or CTRL on a Mac) relate to which links by means of the accesskey attribute.
However, there are problems associated with defining accesskeys on a site as they can over-ride pre-existing keyboard shortcuts in the user’s software.
At one point, it was suggested that using a standard set of numberic keys might be useful and this was, indeed, mandated on UK government sites. However, since the release of Firefox 2.0, there are now significant problems with using this approach as outlined in Gez Lemon’s article Firefox 2.0 and Access Keys. I’ve also made my own feelings fairly plain on the new Firefox implementation of access keys. But, the fact is that, for now, we’re stuck with the current situation and need to bear it in mind when considering the practicalities of implementing accesskeys on a site.
But are accesskeys worth the bother? Do people make use of them?
My own experience and research suggests that many of the user groups that designers assume will want to use accesskeys don’t bother with them. They vary too much from site to site to be really useful. Providing the tab order is intuitive, users prefer to simply tab around pages or use options within their own software (which they know far better than a random site) to jump to specific points on a page or site rather than research a whole list of new keyboard shortcuts on every site they visit.
Gez Lemon and Rich Pedley developed a php AccessKeys class that allows users to define their own access keys . In theory, users could define the same subset of keys on every site that uses this approach. Gez has also since developed an .ASP version.
However, I don’t think either version has been around long enough, or is implemented widely enough, to indicate how many keyboard navigators actually use make use of the facility when it’s offered. I know I’ve never bothered.
Are we just expending a lot of effort to implement an concept that, whilst nice enough in theory, never really caught on amongst users in reality?