W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2006

Re: Key bindings... (user agents - was accesskey was ...)

From: Charles McCathieNevile <chaals@opera.com>
Date: Tue, 10 Jan 2006 19:46:45 +0100
To: geoff@deering.id.au, w3c-wai-ig@w3.org
Cc: "John Foliot - WATS.ca" <foliot@wats.ca>
Message-ID: <op.s26ef7diwxe0ny@pc063.qadoc.oslo.opera.com>

To the extent that key bindings need to be communicated by the user agent  
(which is what ultimately assigns them) to the user, that is covered  
already in UAAG. The ability to note that some waypoint in a site happens  
to be important is something that has to be covered in the relevant  
content specs, it is just that while that markup chosen was reasonable at  
the time, at least HTML 4, SMIL 2.1 got their ideas about implementation  
horribly wrong. Those specifications make claims about platform-based  
implementation that are simply not true, but the model implied was  
followed by the implementations in IE and Netscape/Mozilla, which have  
proven clearly just how wrong they are.

There are various alternative implementations (iCab, JuicyStudio, Opera,  
UBAccess, all spring to mind) that demonstrate that it is possible to make  
something useful of the existing markup.

So, IMHO, WCAG is correct to suggest using them, people who have broken  
implementations have the ability to fix them using the work of people like  
Gez at JuicyStudio or Lisa at UBAccess, manufacturers whose  
implementations more or less work like Opera and iCab should document them  
better, and manufacturers like Mozilla and Microsoft should do something  
about their implementations as well as documentation.

And the HTML and similar groups should leave that bit of markup alone, fix  
their specifications, and let us move on to deal with other problems. (And  
my good mate John should see the light and stop worrying about the key  
attribute so much, in order to concentrate on getting rel right. But he'll  
keep on with both, I guess ;)



On Tue, 10 Jan 2006 19:11:15 +0100, Geoff Deering <geoff@deering.id.au>  

> It seems to me that this needs to be moved to the User Agent  
> Accessibility Guidelines and out of WCAG.  If the whole thing is better  
> managed by the user agent and the client, then that is where it  
> belongs.  Once it is there, developers can be informed in WCAG to avoid  
> setting any key bindings themselves and allow the user agent and client  
> to manage it.  You'd have to try moving it out of markup too.

Charles McCathieNevile                     chaals@opera.com
   hablo español  -  je parle français  -  jeg lærer norsk
      Peek into the kitchen: http://snapshot.opera.com/
Received on Tuesday, 10 January 2006 18:47:03 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:33 UTC