RE: Erratum proposal (was: Re: Avoid deprecated features query)

I have always interpreted WCAG usage of the term "avoid" in the "when
possible but not absolutely" sense. In certain ways, the stricter
interpretation of 11.2 makes my life easier, and it emphasizes the
importance of moving to newer (accessible) technologies. So I kind of
support the clarification / reinterpretation, whichever it is. 

But a flag comes up - a lot of the deprecated features in HTML are
deprecated in favor of CSS. One reason I used to take the softer
interpretation was that CSS support was so shaky you were better off using
deprecated HTML than taking your chances with CSS. That doesn't seem to be
as much of a problem now, but a side effect of the stricter interpretation
would be that we would be practically mandating CSS for most site designs,
which we also encourage in 3.3. I don't think that's a bad thing but I think
a lot of content authors are used to using HTML in certain ways. I see a lot
of pages that use CSS to do a lot of things, yet they still use the <font>
tag, out of momentum as much as anything else as far as I can tell. Although
it shouldn't be much more difficult to use CSS (easier if you structure your
code well), a lot of people aren't going to feel that way. They may feel
that the bar for P2 compliance has been pushed beyond their reach.

I'm not necessarily arguing against this erratum proposal. I just want to
make sure the flag is visible in the decision process. If I were making the
decision, I would probably issue a clarification that "avoid" has the
softer, "when possible" meaning for now, but in WCAG 2.0 go to the stricter
meaning. But if the decision goes the other way I won't oppose it.

Michael

> -----Original Message-----
> From: Wendy A Chisholm [mailto:wendy@w3.org]
> Sent: Wednesday, June 04, 2003 9:43 PM
> To: Al Gilman; w3c-wai-gl@w3.org
> Cc: Pam Galloway
> Subject: Erratum proposal (was: Re: Avoid deprecated features query)
> 
> 
> 
> I found the old issues lists, went through some old minutes, 
> and found 
> contradictory interpretations of "Avoid deprecated features."
> 
> I propose the following erratum to WCAG 1.0:
> 10. Avoid deprecated features
> Added: DD MM YYYY.
> Type: Clarification
> Refers to: Checkpoint 11.2 of 5 May 1999 version
> Description (and correction). The checkpoint says, "Avoid deprecated 
> features of W3C technologies" but it is not clear if "avoid" 
> means "do not 
> use at all" or "use only when absolutely necessary."  Since this is a 
> Priority 2 checkpoint, the stricter interpretation applies 
> and it should be 
> read as, "Do not use deprecated features of W3C 
> technologies."  Deprecated 
> features may be used in a site that claims conformance at 
> Level A.  Level 
> AA requires more work, including replacing deprecated 
> features with updated 
> technology.
> ---
> 
> The other checkpoints that use the word "avoid" are also 
> Prioirty 2 except: 
> "7.1 Until user agents allow users to control flickering, 
> avoid causing the 
> screen to flicker. [Priority 1]"  There are a variety of 
> issues with this 
> checkpoint ("until user agents...," how to test for flicker, 
> what do we 
> mean by "avoid.").  This is a separate issue and should not 
> impact the 
> clarification of 11.2.
> 
> ---
> trail of minutes and issues lists:
> 
> At the 13 April 1999 meeting there was discussion about what 
> we meant by 
> "avoid" but there was no resolution.
> http://lists.w3.org/Archives/Public/w3c-wai-gl/1999AprJun/0015.html
> 
> Greg Lowney commented about an earlier draft and Ian 
> responded that BLINK 
> and MARQUEE "should not be used."
> http://lists.w3.org/Archives/Public/w3c-wai-gl/1999JanMar/0470.html
> 
> In an old issues list for WCAG 1.0, there are still some open 
> issues.  One 
> is related to this issue and it says, "We need to share our 
> interpretation 
> that some use of deprecated is o.k. The CG feels we should 
> either address 
> this in errata or the next version."  However, I am unable to 
> find support 
> that this is the consensus.
> http://www.w3.org/WAI/GL/wai-gl-issues.html#g11
> 
> 13 June 1999, Ian made the following proposal:
> Change wording of 11.2 to read "Avoid deprecated features of
>        W3C technologies. Note: Authors should stop using deprecated
>        features when they are supported by by other technologies such
>        as style sheets."
> http://lists.w3.org/Archives/Public/w3c-wai-gl/1999AprJun/0191.html
> 
> I remember quite a lot of discussion about "avoid tables for 
> layout" and 
> found some evidence of us changing from that to the current 
> wording, "5.3 
> Do not use tables for layout unless the table makes sense 
> when linearized. 
> Otherwise, if the table does not make sense, provide an alternative 
> equivalent (which may be a linearized version). [Priority 2]"
> http://lists.w3.org/Archives/Public/w3c-wai-gl/1999AprJun/0029.html
> 
> This wording comes from the 17 March 1999 meeting:
> http://lists.w3.org/Archives/Public/w3c-wai-gl/1999JanMar/0466.html
> 
> History of issues lists:
> 1. WAI Page Authoring Guidelines Issues List for the Working 
> Drafts (Pre-Rec)
> (last modified Nov 2000)
> http://www.w3.org/WAI/GL/wai-wd-issues
> 
> 2. Summary of Last Call Issues
> http://www.w3.org/WAI/GL/wcgl-last-call-linear
> 
> 3. WAI Web Content Accessibility Guidelines Issues List - Proposed 
> Recommendation
> http://www.w3.org/WAI/GL/wai-pr-issues#layout-tables
> At PR, "Avoid tables for layout" was replaced with a 
> clarified checkpoint:
> 5.3 Do not use tables for layout unless the table makes sense when 
> linearized. Otherwise, if the table does not make sense, provide an 
> alternative equivalent (which may be a linearized version). 
> [Priority 2]
> 
> At 05:07 PM 6/4/2003, Al Gilman wrote:
> 
> >At 03:04 PM 2003-06-04, Wendy A Chisholm wrote:
> >
> >>Hello Pam,
> >>
> >>This is an interesting question.
> >>
> >>Personally, I interpret 11.2 to mean "don't use deprecated 
> features" and 
> >>if you use deprecated features then you don't conform.
> >
> >My recollection would be that we said 'avoid' because we did 
> not feel it was
> >reasonable to expect content completely to satisfy "do not 
> use."  That
> >"do not use" would be interpreted as MUST NOT use in the RFC 
> 2119 sense, 
> >whereas
> >"avoid" indicated a SHOULD NOT in the same terms.
> >
> >If we go back and trace the discussion on layout tables, I 
> believe this
> >interpretation should be borne out.  Not so clear we thought 
> it through
> >in as much depth on deprecated features.
> >
> >I haven't found conclusive proof, but at least at the April 
> 13 meeting the
> >proposal to replace 'avoid' with the simpler "do not use" 
> was considered.
> >The sense that 'avoid' indicates more room for discretion 
> was raised, and
> >the proposed change failed to achieve a consensus either way.
> >
> >  Minutes from 13 April Working Group Teleconference
> >  http://www.w3.org/2002/02/mid/3714876A.B1FE613@w3.org
> >
> >Is there some way to trace from here through the change log 
> for the issues 
> >list
> >to see if this issue were ever resolved at some point?
> >
> >My answer to Pam would be threefold:
> >
> ><draft
> >class="clarification personal">
> >
> >a) yes, the language is meant to leave some room for discretion, for
> >deprecated features to be used where the option is [too difficult].
> >
> >b) this is not a free ride; this means that you have to make 
> a good faith
> >attempt to avoid the use of these features before claiming 
> compliance.
> >Whether you have indeed made a good faith effort is 
> something that would
> >require examination of the facts of the pages, and cannot be 
> determined
> >simply by a clarification of the checkpoint.
> >
> >c) the Working Group will attempt to clarify or interpret 
> the guidelines
> >but cannot take on case-by-case reviews including the facts 
> of a particular
> >site.
> >
> ></draft>
> >
> >Al
> >
> >>It sounds like you have updated some pages but not all.  
> What if you 
> >>clarify the scope of your conformance claim?  For example, 
> "Pages created 
> >>or updated after 1 June 2003 conform at Level AA to WCAG 
> 1.0.  Pages 
> >>created before 31 May 2003 conform at Level A to WCAG 1.0 
> because they 
> >>use deprecated features of HTML 4.01."
> >>
> >>I'm interested in reactions from others on the list.
> >>--wendy
> >>
> >>At 06:42 AM 6/4/2003, Pam Galloway wrote:
> >>
> >>>All
> >>>
> >>>Can you clarify what avoid means in this instance, please.
> >>>
> >>>If a web page has includes some deprecated features and is 
> submitted as 
> >>>4.01 Transistional, which allows for deprecated features, 
> will it be 
> >>>accessible to Priority Level 2 - Guideline 11.2 - Avoid deprecated 
> >>>features of W3C technologies?
> >>>
> >>>I took it that avoid means try not to use, not don't use. 
> I need an 
> >>>answer as we are looking at a site with 100's of pages 
> that needs to be 
> >>>compliant to level 2. Some deprecated features have been 
> dropped and 
> >>>replaced using CSS. Others would take a while to do as the 
> number of 
> >>>pages is extensive.
> >>>
> >>>cheers
> >>>
> >>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
> >>>
> >>>Pam Galloway
> >>
> >>--
> >>wendy a chisholm
> >>world wide web consortium
> >>web accessibility initiative
> >>http://www.w3.org/WAI/
> >>/--
> >
> >--
> >wendy a chisholm
> >world wide web consortium
> >web accessibility initiative
> >http://www.w3.org/WAI/
> >/-- 
>

Received on Wednesday, 4 June 2003 23:07:35 UTC