W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > January to March 2004

Re: REF: 4.1 Move deprecated features

From: Wendy A Chisholm <wendy@w3.org>
Date: Mon, 12 Jan 2004 21:34:45 -0500
Message-Id: <>
To: Don McCunn <Don@designenterprisesofsf.com>, w3c-wai-gl@w3.org


We are aware of our responsibility (refer to "Policies Relating to Web 
Accessibility" which lists policies of 14 countries, the European Union, 
and the United Kingdom [0]).  Creating unambiguous (testable) criteria is 
the key to creating accessible Web content.  If we create a set of testable 
criteria that developers, users, and policy makers (from around the world) 
agree with (using the W3C consensus process), we will be one giant leap 
closer to standards harmonization and accelerating down the path towards an 
accessible Web.

The use of the word "avoid" is a known issue.  I can't find a record of it, 
but we've decided that we need to "avoid using the word 'avoid'." 
<grin>  'Avoid' is ambiguous and we are tasked with writing testable 
(non-ambiguous) criteria (refer to the 2nd requirement in "Requirements for 
WCAG 2.0" [2]).  In other words, we will produce testable criteria, we just 
haven't done it for this one, *yet*.  Deprecated elements and attributes 
are primarily an issue for HTML and in the Working Draft HTML Techniques 
for WCAG 2.0 we make specific recommendations for some non-standard 
elements and attributes.  For example, "Use the embed element within the 
object element for backward compatibility (to embed multimedia objects)." [3]

We are likely to use the word "features" (or something similar) rather than 
"elements and attributes" in the Guideline since we intend this to be 
applicable across technologies.  We can define what we mean by "feature" in 
the techniques documents.  For example, "features" of ECMAScript include 
objects, properties, and methods while in HTML they include elements and 

With respect to the use of "avoid" in WCAG 1.0 (and giving evaluation tool 
developers a common interpretation of this particular checkpoint), I 
proposed wording for an erratum in June [4].  The gist of the proposal is 
that "avoid" should be interpreted as "do not use."  However, as the 
mailing list discussion showed, this is not realistic and I plan to rewrite 
the proposal to say it should be interpreted as "should not 
use."  Obviously, some developers have interpreted it as "do not use" while 
others interpret it as "should not use."  We need to come to consensus and 
issue an erratum to clarify the checkpoint, as well as clarify use of 
specific elements and attributes in HTML Techniques.

This is related to the possibility of publishing a revised WCAG 1.0 that 
incorporates existing errata and addresses some of the ambiguities (I am 
not sure we can address them all, since we want to limit the amount of time 
we spend on WCAG 1.0 so we do not jeopardize progress on WCAG 2.0).  Refer 
to the DRAFT "Tables of WCAG 1.0 Errata" that is being prepared for a 
Working Group discussion of a revised WCAG 1.0 [5].


[0] <http://www.w3.org/WAI/References/Policy>
[1]  <http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=500>
[2] <http://www.w3.org/TR/wcag2-req/>
[3] <http://www.w3.org/TR/2003/WD-WCAG20-HTML-TECHS-20031209/#embed>
[4] <http://lists.w3.org/Archives/Public/w3c-wai-gl/2003AprJun/0262.html>
[5] <http://www.w3.org/2003/12/wcag10-errata-table.html>

At 03:05 PM 1/12/2004, Don McCunn wrote:

>List and Wendy
>I've done some research about how "avoid deprecated features" is 
>interpreted by validation services for WCAG 1.0, Guideline 11.2 with the 
>following results.
>1. AccMonitor by HiSoftware as used on Cynthia Says will apparently not 
>pass any deprecated elements or attributes. This means only strict HTML 
>will pass Prioirty 2.
>2. The Bobby validation test will not pass any deprecated elements of HTML 
>4.0. The Bobby test ignores deprecated attributes. This is clearly 
>indicated in their help documentation.
>3. A-Prompt from the University of Toronto apparently ignores 11.2.
>How does a Web Site indicate to a validation service and the public that a 
>web page is using markup "violated for backward compatibility"? The word 
>"avoid" allows for human interpretation that does not work for software. 
>When you say "avoid deprecated features" do you mean "avoid deprecated 
>elements" or "avoid deprecated elements and attributes"?
>In the book "Accessible Web Sites" Cynthia Waddell indicates that 
>countries are starting to use the WCAG as a part of their laws. This 
>places an awesome responsibility on the WAI to avoid ambiguous meanings. 
>Can the WAI provide a Validation test to avoid the different 
>interpretations that currently exist? If not maybe you could change 
>"avoid" to "minimize the use of deprecated features" or "minimize the use 
>of deprecated elements and attributes." This would clearly signal that 
>human judgement is involved instead of the absolute test that Validation 
>software imposes. Another alternative would be simply leave 4.1d off and 
>just let 4.1a carry the burden.
>Good Luck on working out WCAG 2.0.
>Don McCunn

wendy a chisholm
world wide web consortium
web accessibility initiative
Received on Monday, 12 January 2004 21:35:18 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 21:07:32 UTC