W3C home > Mailing lists > Public > public-html@w3.org > March 2011

Re: Working Group Decision on ISSUE-129 aria-mapping

From: Sam Ruby <rubys@intertwingly.net>
Date: Tue, 01 Mar 2011 06:23:58 -0500
Message-ID: <4D6CD74E.2090908@intertwingly.net>
To: Steve Faulkner <faulkner.steve@gmail.com>
CC: HTMLWG WG <public-html@w3.org>, Richard Schwerdtfeger <schwer@us.ibm.com>, Cynthia Shelly <cyns@microsoft.com>
On 03/01/2011 04:31 AM, Steve Faulkner wrote:
> Thanks to the cahirs for providing decisions on the issues.
>
> comments/questions:
>
> On hgroup; currently the spec says that the hrgoup must have a
>
> "|heading| role, with the |aria-level| property set to the element's
> outline depth <http://dev.w3.org/html5/spec/semantics.html#outline-depth>"
>
> While headings inside hgroup are can have ANY role.
>
> the chairs decision disallows:
>
> " Any changes to how <hgroup> elements are to be interpreted, or how
>     headings contained within such an <hgroup> are to be interpreted."
>
> Which means what is currently in the spec stays.
>
> How does that accord with the Chairs statement in regards to hgroup:
>
> "As such, we find that there is no consensus as of yet as to
> what this element means, and don't wish for this decision to preclude
> any possibility as of yet."
>
> If there is no consensus on the current spec text, will it require that
> this is resolved prior to last call?

We did discuss this.  Short answer: we treated it as "no proposal was 
made in the time allowed; this portion of the change was closed without 
prejudice".

Longer answer:

In the context of the W3C consensus doesn't mean unanimity.  As an 
example, there clearly is disagreement over the default role for img 
elements, but as objections were made that were not countered despite 
everybody being given ample opportunity to do so, we included that 
change in the decision.

In the case of the proposes hgroup change, there is a dispute over what 
the intended semantics of this element is.  Instead of making the case 
for what the intended semantics of this element should be, what we see 
is a concrete proposal with the unsupported assertion that it "better 
matches the semantics that hgroup is intended to convey".

As the proposal lacked this important aspect, we decided to encourage 
the discussion to continue via bug reports.  If amicable consensus can 
be reached: great.  Otherwise, you are correct that we will not require 
bug reports entered at this time to be resolved before proceeding to 
Last Call.

I'll add as a personal note that I wish it were otherwise.  I encourage 
people to submit full and complete Change Proposals.  I believe that we 
have given everybody ample time to do so.  Additionally, I believe that 
this proposal could have benefited by being addressed as a series of 
smaller proposals, something that was encouraged multiple times. 
However, as there seemed to be an insistence to pursue it as a single 
proposal we allowed it to proceed in such a manner.

> If there is no consensus why leave the current text as is, where
> implementors may well assume that hgroup is to be mapped as specified?

If a bug results in an issue, the list of open issues will be forwarded 
along with the Last Call draft.  I'll add that the purpose of Last Call 
is to solicit input, not to state "this is final".

> Just to clarify, from my reading of the decision thes are the changes
> that need to be applied to the HTML5 spec:
>
>     * *button element, input type="image", input type=button*: allowed
>       roles: button, link, menuitem, menuitemcheckbox,
>       menuitemradio,presentation, radio
>     * *h1 to h6 element that does have an hgroup ancestor*: no change
>     * *hgroup element*: no change
>     * *a element that represents a hyperlink* allowed roles: button,
>       checkbox, link, menuitem, menuitemcheckbox, menuitemradio,
>       presentation, tab, or treeitem.
>     * *img element that does not have an empty alt attribute*: default
>       role of img. allowed roles: any
>     * *H1 to H6* allowed roles:  link, menuitem, menuitemcheckbox,
>       menuitemradio,  presentation, tab or treeitem.

While I haven't carefully double checked this list, it does match my 
recollection.

> Regards
> Stevef

- Sam Ruby
Received on Tuesday, 1 March 2011 11:24:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:17:23 GMT