Re: ACTION-128: Draft @summary voting text in conjunction with PF

In a vast oversimplification of this issue, it might be pointed out that a
table *depicts structure* whereas the @summary="" attribute *describes* it.

Those of us who grew up blindless (usually called "sighted") learn to infer
structure from various graphical events before we learn to talk, *without
even realizing we have done so*. For us a description of structure is an
almost DUH! event, but for someone who never "saw" this is not the case. The
attempts to verbally describe structure are akin to explaining to someone
how to put on a jacket without demonstration - it can be done but it's a
clumsy undertaking. [that may be the 150 words in support of option 1]

The purpose of @summary="" is similar to that of @alt in that it is an
attempt (often futile) to capture in text what is for the vast bulk of our
diversitocracy, unnecessary verbiage concerning something that we can
recognize at a glance. Blind people in general don't get to "glance". For
this reason, anything revealing structure is inescapably valuable, including
the much-maligned longdesc.

What I'm saying is that the summary's functional effectiveness must not be
judged by how many UAs or ATs support it, but by whether it might help those
who need it. And we don't got to show no stinking badges (or examples)
because until the need arises (i.e., the table is pretty complex) we don't
know if/how it will be used. Any "substitute" for the current formation will
suffer all these same "shortcomings" and issue resolution will take forever.

I don't know if I have a "vote" but Janina has my proxy if I do.


On Wed, Jul 8, 2009 at 10:31 AM, Ian Hickson <> wrote:

> Joshue,
> Given the recent changes to the spec, two of the options I previously
> suggested for the vote are now equivalent, so I propose we change the
> options we propose to the working group this Thursday to:
>   What do you think HTML5 should say about making complex data tables
>   more accessible?
>    ( ) HTML5 should encourage authors to use the summary="" attribute
>        that was introduced in HTML4.
>    ( ) The text currently in the HTML5 spec, discouraging summary="" but
>        encouraging other mechanisms for describing tables, is fine.
>    ( ) Something else, described below.
>   If you said "something else", please describe it here: [_____]
> I think if we have these options we should definitely also include a
> 150-word essay for each of the first two options, written by proponents of
> those options, that argue each case, so that people who aren't familiar
> with the issues can read up on them before voting.
> What do you think? Is this what we should propose to the working group?
> Sam informs me that we're supposed to decide this by tomorrow (Thursday),
> for presentation to the group by the conference call.
> --
> Ian Hickson               U+1047E                )\._.,--....,'``.    fL
>       U+263A                /,   _.. \   _\  ;`._ ,.
> Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'


Received on Wednesday, 8 July 2009 09:06:02 UTC