W3C home > Mailing lists > Public > public-html@w3.org > June 2009

Re: PF Response: @Summary

From: David Poehlman <poehlman1@comcast.net>
Date: Thu, 4 Jun 2009 04:18:26 -0400
Cc: Janina Sajka <janina@rednote.net>, Sam Ruby <rubys@intertwingly.net>, Chris Wilson <Chris.Wilson@microsoft.com>, Michael Cooper <cooper@w3.org>, Dan Connolly <connolly@w3.org>, Mike Smith <mike@w3.org>, W3C WAI Protocols & Formats <w3c-wai-pf@w3.org>, Gez Lemon <gez.lemon@gmail.com>, Laura Carlson <laura.lee.carlson@gmail.com>, "wai-liaison@w3.org" <wai-liaison@w3.org>, John Foliot <jfoliot@stanford.edu>, www-archive <www-archive@w3.org>, public-html@w3.org
Message-Id: <45E132E4-D11B-420E-8951-CE71A732B804@comcast.net>
To: Ian Hickson <ian@hixie.ch>
"I believe that restoring the summary="" attribute to the HTML5 language
would negatively impact users, which is a greater concern to me than any
possible impact to the specification itself."

Rushing this does more harm than good in any event.  The <caption> tag  
already fills a need.  how are we to distinguish when it is used as  
table summary and table caption?  I agree with those who have said  
that using government is a red herring simply because it is fluid and  
all over the map.  We could site several sources and they'd all be  
conflicting.  The summary attribute in and of its self is not  
harmful.  I've asked before that since <p> blah blah blah</p> is so  
poorly used, should we take it out of the speck or any other tag  
thusly abused?

What troubles me as with some other accessibility attributes is that  
the information in summary is "hidden" but the reason for this is  
understandable for if used correctly, it provides information that is  
not useful to anyone but a search engine or an assistive technology  
user; except where someone not using assistive technology has a  
cognative need.  in this event, user agents could be asked to display  

We need more research on why summary came into being in the first  
place and more places where it is used as intended, meaning that a  
huge complex data stream captured in a table is adaquately narratized  
in the summary.  It is not enough to say: "this table contains 1,000  
rows and 600 columns and shows temperatures across the middle united  
states for the past 20 years".  The summary needs to go on to say  
something about the trends depicted in the stream and the ranges and  
much much more in order for it to be useful.  We may be forgetting  
that there are still a lot of assistive technology users who are not  
able to examine a table in other than liniar form even if the  
technology is capable of providing this information either due to poor  
documentation or poor results when attempting to do so.  We should  
also not forget all the colored lines running through the table that  
make things still more interesting.

"look at the graph in the upper left quadrant of the table and you  
will see that where the green and blue lines intersect is the same as  
where the red and orange lines of the graph in the lower right  
quadrant intersect".

This can Be much better understood when summarized as: "The average  
temperatures for the beginning and end of the period are roughly the  
same" or something similar.  Yes, you need to tell me about the data,  
I cannot use assistive technology to examinie it to the point where I  
can get the information that is available at a glance out of it that a  
person with a glance to spare can because in this instance for  
example, it is not possible to follow the colored lines.  It also  
needs to be associated with the table such that as I use my assistive  
technology or user agent, I cn readily access the information  
unambiguously.  <scratching head...now, is that caption telling me  
what I need...</scratching head>.  We really need to be careful to  
provide a one to one relationship between markup and outcome as much  
as possible and not stuff one tag with too much responsibility.

If you think is summary is abused, wait till you start drafting  
caption for this and see how much abuse you get.

On Jun 4, 2009, at 1:19 AM, Ian Hickson wrote:

On Wed, 3 Jun 2009, Janina Sajka wrote:
> We request the table summary tag be restored in HTML 5 as per previous
> communications:
> http://lists.w3.org/Archives/Public/public-html/2008Aug/0213.html
> Rationale:
> Summary serves a need

Is the need not served by <caption>?

> and serves it well.

Evidence has been presented that actually not only does summary="" not
actually in practice serve the need for which it is intended well, but
that it causes extra harm to users, with certain sites including
information in the summary="" attribute that should be available to all
readers universally but that is, due to the use of the summary=""
attribute, in fact not available universally.

Could you elaborate on why you believe summary="" serves its purpose  

> It is familiar to users.

So the data I have seen on the matter suggests that users in fact do not
find the summary="" attribute to be useful. When you say that the
attribute is familiar to users, do you mean that they like it? On what  
you base this assessment?

> It is properly utilized on many web sites which strive to be  
> accessible.

Actually the research I have seen suggests quite the opposite -- that
sites that strive to be accessible actually use it so incorrectly that
they reduce the accessibility of the site to users that don't have (or
need) screen reader software, for example by including important
information in the summary="" attribute that would be better placed in a
table caption or elsewhere on the page.

> *	If it didn't exist, we'd need to invent it. Indeed, such alternative
> *	approaches as have been proposed constitute a "reinvention" of
> *	summary. PF wishes to move on to address accessibility concerns in
> *	areas, such as canvas, where no good solution currently exists  
> rather
> *	than reinventing summary.

The summary="" attribute has failed improve accessibility for tables.  
I do
not believe we should ignore this. We have a moral imperative to make  
Web accessible to all users, and this means being honest about what  
and what does not. The summary="" attribute, according to all the data I
have seen, does not work. Therefore we are IMHO duty-bound to solve the
problem that it attempted to address.

HTML5 attempts to solve the problem using the <caption> element. I
understand that PF may not have the bandwidth to review this proposal,  
I do not believe that we should let PF's limited resources prevent
potential progress in the development of accessible HTML.

> *	We reject the argument that summary should be removed from the HTML
> *	specification because it is not implemented on most web sites.

I do not believe this argument has ever been made. I would also reject
such an argument.

> *     We note
> *	that accessibility is poorly supported on most web sites. The wider
> *	web is not an example of good practice.

Indeed, we should do things that would help address this. The summary=""
attributes contributes to this problem by encouraging authors to  
accessibility to be an extra issue to be dealt with separately rather  
a fundamental concern to be dealt with at first. Using the <caption>
element for the purpose of orienting users of tables brings the  
subject to
the fore, which I believe has a better chance of addressing this very
important issue.

> *	We need summary for backward compatibility.

HTML5 supports implementing the summary="" attribute for backwards
compatibility as currently written.

> *	We note that summary is often used as a technique for accessibility
> *	support where governmental regulations require governmental web  
> sites to be
> *	accessible.

The <caption> element can equally be used as such a technique.

> *     An example is the U.S. Government's Social Security
> *	Administration (SSA) pages as SSA conforms to its "Section 508."
> *	mandate:
> 	http://www.usdoj.gov/crt/508/web.php

This very page is in fact a clear demonstration of the very problem that
summary="" introduces. If you search for the table regarding the
mathematics department, you will see that the screen-reader content is  
fact more useful to the reader than the "ordinary" content -- there is
information in the summary="" attribute that should rightfully be
accessible to all users. The summary="" attribute actually encourages  
creation of pages that are less accessible to non-screen-reader users.
This is especially ironic because it is primarily authors who are
attempting to be as universally accessible as possible who are afflicted
by this problem.

Using the <caption> element in the very same example would result in
content that is accessible to all.

> 	http://www.access-board.gov/sec508/guide/1194.22.htm#(g)

To quote from this document:

| Although highly recommended by some webpage designers as a way of
| summarizing the contents of a table, the "summary" attribute of the
| TABLE tag is not sufficiently supported by major assistive technology
| manufacturers to warrant recommendation. Therefore, web developers who
| are interested in summarizing their tables should consider placing  
| descriptions either adjacent to their tables or in the body of the
| table, using such tags as the CAPTION tag.

This is in fact exactly what HTML5 says.

> *	If summary is removed, U.S. Government web sites, might find it more
> *	difficult to conform to HTML 5.

As far as I can tell this concern is unfounded; the <caption>  
attribute is
in fact encouraged by the very same government (as quoted above) to be
used exactly as HTML5 recommends in a manner consistent with the goals  
the summary="" attribute.

> *     We further note that Section 508
> *	regulations apply to U.S. state and local governments, and that
> *	similar accessibility requirements are emerging in Canada, the U.K.,
> *	the E.U., Australia, and elsewhere.

I entirely agree that accessibility is important; the current text in  
HTML5 specification encouraging the use of <caption> where previous
specifications encouraged summary="" is in fact intended to increase the
accessibility of pages.

> *	Restoring summary in HTML 5 would not, in our understanding,  
> negatively
> *	impact HTML 5 in any way.

I believe that restoring the summary="" attribute to the HTML5 language
would negatively impact users, which is a greater concern to me than any
possible impact to the specification itself.

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

Jonnie Appleseed
with his
Hands-On Technolog(eye)s
reducing technology's disabilities
one byte at a time
Received on Thursday, 4 June 2009 08:19:13 UTC

This archive was generated by hypermail 2.4.0 : Saturday, 9 October 2021 18:44:48 UTC