[whatwg] Accesskey in Web Forms 2

Ian Hickson ian at hixie.ch
Tue Dec 7 19:03:47 PST 2004

On Thu, 11 Nov 2004, Mikko Rantalainen wrote:
> Derek Featherstone wrote:
> > 
> > I still like the XHTML 2 proposal of the access attribute:
> > 
> > For single page:
> > 1. Authors define key access points for items in their documents 
> > (their search form, individual form fields, other forms, whatever)
> > For all pages/sites 1. Authors define key access points across *all* 
> > sites should be defined that mimic or bind to <link rel="" /> elements 
> > and their defined values - both currently existing and expanded: home, 
> > search, help, up, next, prev, privacy, accessibility, copyright, 
> > etc...
> This is exactly what I've been thinking. I don't think that 'rel' is the 
> best choice to bind with but it's a good start.

HTML has allowed the above for links since at least HTML2, and so far to 
my knowledge only one UA has used this (Opera, using rel="next" for its 
Fast Forward feature), and that only for one value (next). I don't see why 
UAs would do any better with access="".

> If the new attribute is called 'access' one could write <input 
> type="text" access="person-first-name"> and UA could provide 
> 'person-first-name' in the list of possible actions. The user could then 
> bind a key to that action if he so decides. An UA running on MS Windows 
> could even add another menu "_A_ccess" and populate it with all access 
> points.

I just don't see that UAs would do this. The trend recently has been to 
minimalist UIs, removing menus, not adding them.

> The idea behind allowing special 'access' attribute is to allow authors 
> to specify objects to mean the same thing even though the actual 
> implementation differs -- one web site could bind 'search' access point 
> to a link going to search page and another could bind the same access 
> point to text entry box.

rel="search" already allows this, to some extent... We could add rel="" to 
the submit buttons, or maybe forms, would that be a good compromise? 
Basically it's taking an existing solution, and expanding its scope a bit, 
without having to introduce a new attribute with all-new semantics, etc.

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

More information about the whatwg mailing list