Re: Step 4.c

Hi all,

 From the Glossary definition:

"at least one of the following is true:

- the conforming version can be reached from the non-conforming page via 
an accessibility-supported mechanism, or"

In this case the non-conforming page MUST at least meet Conformance 
Requirement #5, and therefore the 4 SC that Detlev mentioned. Otherwise, 
the interference would prevent access to the "accessibility supported" 
mechanism mentioned in the above definition.

Thus, when the accessible version is reached from the inaccessible one, 
it should be mandatory to evaluate at least the 4 SC that may lead to 
interferences.

Nevertheless, I guess that interferences may be caused by other barriers 
that are out of the scope of the 4 SC specifically mentioned by WCAG. 
For example, imagine a link to reach the conforming alternate version 
inside a deep ocean of inaccessible content and zillions of other links. 
Someone could argue that the link itself is accessibility supported, but 
the reality is that the inaccessible content is clearly interfering with 
access to the conforming version. Or will we force the user to explore 
the whole page to find the Lost Arc?

For the other two cases (where the non-conforming version can only be 
reached from conforming versions / pages), in principle it seems that 
conforming alternate version would be enough, and that we might not need 
to assess the parallel, non-conforming one. (of course, we need to 
assess the "bridge" pages anyway).


Note 4 of the definition states:

"Note 4: Alternate versions may be provided to accommodate different 
technology environments or user groups. Each version should be as 
conformant as possible. One version would need to be fully conformant in 
order to meet conformance requirement 1."

Which, in my opinion, suggests that the non-conforming version(s) should 
also be tested to assess "how much" of that accessibility level has been 
reached. At least in any scenario that aims to provide real 
accessibility, and not just "technically conform".


In any case, we have also the following text from the "Understanding 
Conforming Alternate Versions":

"A concern when permitting Web pages that do not satisfy the Success 
Criteria is that people with disabilities will encounter these 
non-conforming pages, not be able to access their content, and not be 
able to find the “conforming alternate version." A key part of the 
Alternate Versions provision, therefore, is the ability to find the 
conforming page (the alternate version) from the non-conforming page 
when it is encountered. The conformance requirement that permits 
alternate pages, therefore, also requires a way for users to find the 
accessible version among the alternate versions."

Which, again, reinforces the argument that the evaluation should ensure 
that the mechanism is "findable" (and not just "accessibility 
supported"). If the non-conforming version contains any content that 
interferes, or if the mechanism is "hidden" or hard to find, I think we 
cannot say that the requirements for "conforming alternate version" are met.


Finally, I think that any procedure or statement that refers to 
conforming alternate versions in the Evaluation Methdology should be 
very carefully worded, containing -or pointing to- the following note in 
the Understading document:

"Note that providing an alternate version is a fallback option for 
conformance to WCAG and the preferred method of conformance is to make 
all content directly accessible."

Unfortunately, I am seeing many developers that read the "conforming 
alternate version" term in a similar way to the outdated and harmful 
concept of "text version" or "blind / disabled version", that is, a 
simple way to completely bypass WCAG as a whole, and not as a 
last-resort solution for those special cases where accessibility cannot 
be reached with existing techniques in the "everyone" version.

Note also that the rationale behind allowing conforming alternate 
alternate versions in the Understanding document refers to strong 
tchnical reasons (new technologies that -temporarily- lack accessibility 
support, specific UX aimed at certain disabilities) or legal reasons 
(copyright issues, legal permissions, etc.).

This means, in my opinion, that conforming alterate versions should 
never be promoted as "global solutions" to bypass WCAG. I think that we 
must be very careful and rigurous when we mention this "alternate" 
possibility of the CR #1. Otherwise people could interpret that W3C is 
saying that they are a valid solution for any situation, even for some 
that could be easily addressed with accessibility techniques that are 
already available.

Regards,
Ramón.

PS: Please don't read my words as a complain against developers/owners 
or their intentions. My concerns are more related to possible 
misinterpretations that people might extract from the WCAG terminology.

Alistair wrote:

 > From the WCAG 2.0 conformance requirements for conforming alternate 
versions it states:
 >
 > "Note 4: Alternate versions may be provided to accommodate different 
technology environments or user groups. Each version should be as 
conformant as possible. One version would need to be fully conformant in 
order to meet conformance requirement 1."
 >
 > It does not mention checking non-conformant versions - simply that 
efforts should be made to make them as conformant as possible.  So I 
don't really see why you think "skipping the non-conforming page and 
checking the conforming alternate version is not in line with the WCAG 
conformance requirements".
 >
 > Is there something I have missed, or not read?

Received on Thursday, 22 May 2014 18:03:12 UTC