Re: Access Key (and proposed HTML4 erratum)

Hi folks,

one of the reasons I use iCab a lot is that it shows what accesskeys are
available, visually, and it activates the links in a single keypress - I find
this slightly more convenient than the double action of <accesskey><enter>
(especially when I am reduced to using one hand) because it saves me
travelling across the keyboard (for the same reasons I find this preferable
to <modifier><accesskey><enter> ...)

I understand the issue that leads to people wanting to make accesskey a
two-step process. I suspect this is a minority case, and should be handled by
allowing for configuration in the User Agent, although I think that either
approach is acceptable - the more so if users can readily find out what
accesskeys are available and know that they are not going to override system
functions they rely on.

Accesskey introduces some general problems for User agents - in many systems
certain keys are reserved, or not available (how many people know how to
activate the 'kaf' key?), and the user agent has to find some way of making
the functionality available. In Amaya, the accesskey overrides the system
functionality, which is an obvious problem. In Explorer, certain accesskeys
are simply ignored, which also creates a problem, especially without clear
documentation of which keys will be ignored.

For what it is worth, the WAI Protocols and Formats group is aware of this
issue, and expecting to deal with it in its review of XHTML 2.0. Further
input is of course welcome - for people with member access I recommend
copying the PFWG at w3c-wai-pf@w3.org but for poeple without member access
this thread is a useful place to collect information that will remain
publicly available.

cheers

Charles McCathieNevile

On Fri, 27 Sep 2002, Tantek Çelik wrote:

>
>Steven, Wendy,
>
>Thanks for bringing up and furthering this discussion.
>
>This (the lack of clarity on what an accesskey does on a hyperlink) has been
>an issue for us (Microsoft).
>
>IE/Windows takes the more conservative approach of *not* activating, and
>just focusing the hyperlink.  This allows for better handling of duplicate
>accesskeys for example, e.g. focus to the first element that has the
>accesskey after the currently focussed element, thus you can "cycle" through
>such elements by continuing to press the same accesskey.
>
>IE5/Mac took the more literal approach of what was implied by the HTML4 spec
>of activating the link (as it says in the example).
>
> http://www.w3.org/TR/html401/interact/forms.html#access-keys
>
>================
>In this example, we assign an access key to a link defined by the A element.
>Typing this access key takes the user to another document, in this case, a
>table of contents.
>
>
><P><A accesskey="C"
>      rel="contents"
>      href="http://someplace.com/specification/contents.html">
>    Table of Contents</A>
>================
>
>Since this is only stated in the spec as part of an example, it is
>informative (not normative).  However, there is a definite direction of
>implementation implied by the spec, so while filling out our HTML4
>implementation in IE5/Mac, we took it.
>
>But it has been perhaps three years since I took that interpretation, and
>have realized that not only is it
> a) inconsistent between our products, but also
> b) inconsistent among elements (other elements focus rather than activate)
> c) disadvantageous to handling duplicate accesskeys
> d) disadvantageous for accessibility from the perspective of the
>possibility that a hyperlink is activated by the user accidentally without
>first being able to see (via the transfer of the focus) what hyperlink they
>are activating.
>
>
>Thus I now tend to agree with Masafumi Nakane [3].
>
>Personally I would like to see that section of HTML4 errata'd
>
>from:
>
>=======
>Typing this access key takes the user to another document, in this case, a
>table of contents.
>=======
>
>to:
>
>=======
>Typing this access key focuses a hyperlink to another document, in this
>case, a table of contents.
>=======
>
>
>Since this is merely a change to an informative example, this shouldn't be
>considered a change to the conformance/functionality of HTML4 and hopefully
>can be made easily.
>
>
>Thanks for your consideration.
>
>
>Tantek
>
>
>
>On 9/27/02 10:37 AM, "Wendy A Chisholm" <wendy@w3.org> wrote:
>
>>
>> Steven,
>>
>> Looking on the WAI lists, I see a variety of responses, particularly in the
>> thread started by Aaron Leavanthal on 24 April 2001. [1]
>>
>> For example:
>> Gregory Rosmaita thinks it should be user configurable. [2]
>> Masafumi Nakane thinks we should take a conservative approach and *not*
>> activate. [3]
>> Kelly Ford makes a distinction between a button (activate) and a link
>> (focus). [4]
>>
>> I would ask the UAWG.  They have a variety of test suites for accesskey
>> [5,6] and UA behavior is the primary topic of UAAG.
>>
>> --wendy
>>
>> [1] http://lists.w3.org/Archives/Public/w3c-wai-ig/2001AprJun/0241.html
>> [2] http://lists.w3.org/Archives/Public/w3c-wai-ig/1999OctDec/0725.html
>> [3] http://lists.w3.org/Archives/Public/w3c-wai-ig/2001AprJun/0274.html
>> [4] http://lists.w3.org/Archives/Public/w3c-wai-ig/2001AprJun/0253.html
>> [5] http://www.w3.org/WAI/UA/TS/html401/cp0101/0101-ACCESSKEY.html
>> [6] http://www.w3.org/WAI/UA/TS/html401/
>> At 10:52 AM 9/27/02, Steven Pemberton wrote:
>>
>>> Editing the XHTML 2 spec, I came across something I wasn't sure of.
>>>
>>> If an access key is on a link, should the access key take you to the link,
>>> where you must then activate the link (with return or space or whatever), or
>>> should it immediately follow the link.
>>>
>>> HTML 4 says follow the link, but I note that UAs aren't consistent here. My
>>> personal feeling is that the link should get focus, so that you can then
>>> decide what to do with it, but I'd like to hear from WAI people.
>>>
>>> Thanks!
>>>
>>> Steven Pemberton
>

-- 
Charles McCathieNevile  http://www.w3.org/People/Charles  tel: +61 409 134 136
SWAD-E http://www.w3.org/2001/sw/Europe ------------ WAI http://www.w3.org/WAI
 21 Mitchell street, FOOTSCRAY Vic 3011, Australia  fax(fr): +33 4 92 38 78 22
 W3C, 2004 Route des Lucioles, 06902 Sophia Antipolis Cedex, France

Received on Saturday, 28 September 2002 14:24:49 UTC