Re: [#293] Summary for tables

The only reason to use null summary would be the same as for null alt.  If,
without null summary, something would be rendered that would be as gibberish
than using a null summary would be good.  I'd like to teach developpers the
difference between a data table and a lay out table if they don't know and
this is one of them.  I like your example and find the captions, although my
screen reader does not render them as captions, helpfull as well.  My main
concern is that today, we see increasing use/over use of the summary
attribute when it comes to lay out and this only makes for clutter on the
braille display or in the ear and I suspect in other modalities as well
where it is rendered and announced.

----- Original Message ----- 
From: "John M Slatin" <john_slatin@austin.utexas.edu>
To: "David Poehlman" <poehlman1@comcast.net>; <w3c-wai-gl@w3.org>
Sent: Tuesday, June 10, 2003 11:43 AM
Subject: RE: [#293] Summary for tables


Excellent points about the difference between a data table and an image;
I aboslutely agree that the summary doesn't serve as "equivalent
alternative" for the table, but rather as an explanatory or descriptive
statement to irent the user to the table.

The reason I suggested the possibility of a null summary for both data
and layout tables is that for some relatively straightforward data
tables the <caption> element can provide sufficient information, and it
seems pointless to insist on having a non-empty summary in such cases.
Ideally, of course, the summary and <caption> provide complementary
information.

For example, on the course schedule page at
http://www.cwrl.utexas.edu/~slatin/webaccess/2002/schedule2002.htm, I
used the <caption> element to display the titles for the course units,
while the summary attribute gives the date range for that unit.  The
rationale for this is that sighted users can quickly get the date range
by scanning the page, whereas people using screen readers would
otherwise have to listen to the entire schedule-block to find out where
it ends.



John Slatin, Ph.D.
Director, Institute for Technology & Learning
University of Texas at Austin
FAC 248C
1 University Station G9600
Austin, TX 78712
ph 512-495-4288, f 512-495-4524
email jslatin@mail.utexas.edu
web http://www.ital.utexas.edu



-----Original Message-----
From: David Poehlman [mailto:poehlman1@comcast.net]
Sent: Monday, June 09, 2003 4:03 pm
To: w3c-wai-gl@w3.org
Subject: Re: [#293] Summary for tables



I would like to submit that Although I support the null summary for lay
out tables if the null summary actually works as a null rather than
putting some sort of flag that the screen reader or other rendering
agent will speak/render, Alt and summary are two different things.  In a
data table which is quite complex, it is often a good idea for usability
for a narrative to be provided as part of the table such as in a summary
in order to get the big picture of the table.  This is more like
narative description than alt/replacement for the table since It
provides an overview and possibly some instructive information
concerning the table and the data in it.  A data table is not a
graphical element as such but is a structural element or block of
structural elements holding textual relationships, while though
benefitting vision are not enherent to it since tactually and via use of
a screen reader or other device which allows appropriate navigation, the
relationships among the data and data sets can be discerned.

I propose then that null summary be used for lay out tables as per my
comments above, and that summary be used with appropriate content for
data tables but that the later might be constrained by the complexity of
the structure.  In other words on the second point, some times, there is
nothing to say beyond the rendering of the table so a null summary would
be acceptable.

Thanks for allowing me to intrude.

Received on Tuesday, 10 June 2003 12:01:05 UTC