W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2003

Re: WCAG Checkpoint 5.3

From: Phill Jenkins <pjenkins@us.ibm.com>
Date: Fri, 27 Jun 2003 12:25:21 -0500
To: w3c-wai-ig@w3.org
Message-ID: <OF434344D7.34815D11-ON86256D52.005BDDA5-86256D52.005FC40F@pok.ibm.com>


Please be careful when quoting checkpoints and try to include the URL to
the checkpoint you're quoting.

In WCAG 1.0 Checkpoint 3.3 [see note 1] says it's a priority 2 to use style
sheets, it doesn't say - must.
In my view it is not in conflict with WCAG 1.0 checkpoint 5.3 [see note 2]
which says do not use tables for layout, unless they makes sense
linearized.  Many web sites do use tables for layout because of backwards
compatibility problems with browsers and ATs.

However, the argument should not about using table or CSS - the problem
occurs with either technique when the content doesn't make sense when
linearized.  Screen readers render the content in a serial fashion, so
whether using tables or CSS to make layout, the reading order needs to make
sense.  Typically reading order is not as critical for layout tables or CSS
positioning as it is in data tables because there are no column or row
headings to worry about.

But I agree that there appears to be a conflict because folks interpret 3.3
and 5.3 and add words like must, only, and always, which were cleverly
omitted from the checkpoint wording to reach consensus (my opinion cause I
was there).  For example, 3.3 neither says "must use CSS" to achieve
priority 2, nor does it say "only use CSS",  nor does it say "always use
CSS".  There is also a note in checkpoint 5.3 (see note 2) that is part of
the normative wording of the checkpoints: "Note: Once user agents support
style sheet positioning, tables should not be used for layout." Notice the
missing words of only, always, and must, but the presence of the word
should.

So I again conclude that the accessibility requirement here is not about
whether you are using CSS or tables as much as it should be about making
the reading order make sense - something missing from WCAG 1.0 checkpoint
3.3 - something that should be considered for an ERRATA for  WCAG 1.0 so
that these debates can stop.


Note 1 WCAG 1.0 Checkpoint 3.3
      http://www.w3.org/TR/WCAG10/wai-pageauth.html#tech-style-sheetst
Note 2 WCAG 1.0 Checkpoint 5.3

http://www.w3.org/TR/WCAG10/wai-pageauth.html#tech-avoid-table-for-layou


Also, when reading the 24 June 2003 public draft of WCAG 2.0, this
principle seems to be somewhat buried to me.  Extended Checkpoint 3.4 (see
note 3) talks about consistent and predictable layout, but not
understandable layout. But Core Checkpoint 1.3 (see note 4) does attempt to
address the concept in the informative example 1 about multiple columns of
information.  However, the point of reading order making sense when
presentation markup is removed is not explicitly addressed - a comment for
the working group meeting next week..

Note 3 WCAG 2.0 public draft 24 June Checkpoint 3.4
      http://www.w3.org/TR/WCAG20/#consistent-behavior
Note 4 WCAG 2.0 public draft 24 June Checkpoint 1.3
      http://www.w3.org/TR/WCAG20/#content-structure-separation

Regards,
Phill Jenkins
IBM Research Division - Accessibility Center
11501 Burnet Rd,  Austin TX  78758    http://www.ibm.com/able
Received on Friday, 27 June 2003 13:25:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:14:10 GMT