From W3C Wiki
< PF‎ | XTech‎ | HTML5‎ | Table
Jump to: navigation, search

headers Attribute for TABLE: Problem Description

Current Assistive Technologies (ATs) do not support the HTML5 way of associating header cells and data cells. But they do support id/headers.

This means that complex irregular tables cannot be marked up in HTML5 to provide current asssistive technologies with the required information about relationships between headers and data.

So in order to support Assistive Technologies that do not implement the HTML5 algorithms, authors who want to use HTML5 will be forced to write non-conformant code.

Stage 1: HTML WG Issue 72

Joshue O Connor, Gez Lemon and Laura Carlson have completed the deliverable for HTML WG Action 72.

Joshue, Gez and Laura requested that the definition of the headers attribute in the specification be extended to allow it to reference a TD. This would make it possible for complex data tables to be marked up accessibly.

The headers/id markup is functional and works today. Results of some recent testing of @headers is available.

Current Status (2008-08-31)

At the 28 August 2008 HTML WG issue tracking telecon, the issue of headers not being allowed to reference a td in the current version of spec was discussed.

It was decided by the co-Chair, Chris Wilson, after discussion and consideration of the pros and cons that the current spec should be changed to allow headers to reference a td element. Chris "took action item assigned to josh and will figure out how to ensure edit gets made in timely f ashion" source: HTML WG Telecon Minutes 2008-08-28

The associated action item (HTML WG Action 72): "rewrite spec to reinstate id/headers AND their functionality by specifically stating that headers are allowed to reference a td. Reword the current definition of the headers attribute so that each of the space separated tokens must have the value of the ID value of a th or td element."

This means that the HTML5 spec will contain a method to mark up complex and irregular data tables accessibly that is well supported by current assistive technologies.

It is also a positive sign that the W3C process is working within the HTML WG.

A few further points in regards to this matter:

  1. Allowing a header to reference a td does not preclude the use of an algorithm which in the majority of cases may obviate the need for the headers attribute. It merely recognises a mechanism that can provide the required associations in cases where the algorithm cannot.
  2. The issue of headers on td arose from a bug post. It was escalated to the Chairs of the HTML WG by the editor "Reassigning to Mike for arbitration." It has been decided by Chris Wilson as Chair that a header should be allowed to reference a td.

So the question is whether the editor will abide by an arbitration process that he invoked, or because he disagrees with the decision choose to ignore it?

He has recognised the authority of the chairs to override his authority:

    Sure, so long as I am editor, the chairs have the authority (and responsibility) to override me if I make mistakes." (source: Ian Hickson, post to public-html)

In the case of headers it would seem that the Chairs have decided the editor has made a mistake in not allowing it.

But it now appears that, despite the editor sending the issue to arbitration, and publicly stating that the chairs have authority over what goes into the spec, the editor will now ignore the result of the process he himself set in motion:

    <Hixie> fwiw my plan with this tables stuff is to do exactly what i was going to do anyway, which is to get to it when i get to it, and when i get to it, read all the arguments and research, and come to a conclusion based on that" source: Ian Hickson, WHAT WG IRC Log (2008-08-29)


Email Threads

Thread 1: table headers - clear description of problem

Thread 2: Deliverable for Action 72 @headers

Thread 3: @headers issue resolved - allowing a td to be referenced by a header to be in the HTMl5 spec