Re: "accesskey" - Techniques open issue

I believe there are a lot of issues with accesskey. For a start, only some
users wll have the key specified available to them. And as Gregory says,
using it may take over a key function for a particular user. An intelligent
user agent would prevent this happening by remapping the key to something
else, but then it is not clear how a user could identify the accesskey. It is
also unclear how different systems are expected to use it - some software has
ordinary keys as hotkeys (lynx, Opera), some uses accelerator keys (alt in
windows, ctrl or alt or meta in unix, ctrl and command keys in macintosh) -
what should the author tell the user, without knowing how the system
works? At the moment the only implementations I know of are i Internet
Explorer, but they are not consistent between versions.

I recommend that we note the potential for conflict with existing bindings
means that the User Agent should handle that and notification of how to
activate keys (and register that as a dependency on the User Agent Group) and
that we ask the PF group to work with the HTML group to respecify accesskey
in such a way that it is not dependent on a particular keyboard arrangement.

Charles McCN

On Thu, 8 Jun 2000, Wendy A Chisholm wrote:

  Hello,
  
  As we were wrapping up the Techniques document for WCAG 1.0 we resolved 
  that, "There most likely will be usability issues [with accesskey], but we 
  will not be able to cover them in this initial release of the document. 
  Therefore, this remains open to be discussed for the next version."