- From: Phill Jenkins <pjenkins@us.ibm.com>
- Date: Fri, 23 Sep 2016 08:27:58 -0500
- To: Jonathan Avila <jon.avila@ssbbartgroup.com>
- Cc: "w3c-wai-ig@w3.org" <w3c-wai-ig@w3.org>
- Message-Id: <OFD93343DD.C2ED9827-ON86258037.0046C6BF-86258037.0049F9F1@notes.na.collabserv.c>
Jonathan write:
"I do support computer learning to solve issues in a pinch and provide a
bridge or fix for inaccessible content. However, I am hesitant to rely
off of these cloud based solutions as the on demand approach for permanent
accessibility. Cloud based services aren?t reachable everywhere nor in
all situations.
Agree for today, and for tomorrow we'll need more reliability and
backup.
. . . While computers may be able to provide you good information about
particular images they may not be able to provide an equivalent to image
without using contextual clues from the text.
Contextual clues for surrounding text is part of the solution (or
should be), so is learning from getting it right the last time, etc. the
solution includes all the parts of image recognition, machine learning,
human language understanding, etc.
While Watson and others may be able to provide this, accessibility
shouldn?t rely on paid services that the user doesn?t have control over.
I was thinking more that the site would pay, and/or be included in
the browser and/or AT, but agree the user needs control and preferences.
...If authoring tools generated accessible content and frameworks had
built in accessibility it would mitigate the effort on authors. I agree
there is more user agent manufacturers need to and can do.
Agree, I really stress the need for accessible frameworks, pattern
libraries, accessible style guides, and inclusive designs designed from
those resources, but as a community we seem to spend too much time &
effort relatively on the wrong audience - developers and testers.
However, without markup it may not always be possible to know the
intention of the author. So creating specifications that allow authors to
communicate their intentions with markup rather than using generic markup
is beneficial.
Agree, even with markup we don't always know the intent of the
designer and developer, especially when the mark-up is wrong when they use
poor design methodologies, poor authoring tools, inaccessible frameworks,
etc.
I think as web development both moves to more continuous delivery
approach from Agile and also as web development matures with old software
concepts like hot on-demand patches, those on-demand technologies can
actually help fix the source so end users the next day get the benefits of
the now built-in accessibility. Timing is everything. My point is that
we should push in that direction.
. . . ? why wouldn?t we provide something formatted in a way that is
adaptable for users with disabilities as well?
totally agree, adaptability is key. WCAG Guideline 1.3 Adaptable
. And with the aging population percentage growing, we'll get more demand
for better authoring tools, frameworks, and built-in browser adaptability
- and so lets all join in and push more in that direction.
___________
Regards,
Phill Jenkins,
Senior Engineer & Business Development Executive
IBM Research - IBM Accessibility
ibm.com/able
facebook.com/IBMAccessibility
twitter.com/IBMAccess
ageandability.com
From: Jonathan Avila <jon.avila@ssbbartgroup.com>
To: "w3c-wai-ig@w3.org" <w3c-wai-ig@w3.org>
Date: 09/22/2016 08:04 PM
Subject: RE: WCAG 2.1
Ø In other words, no need to wait for the content author to make the
repair
and add the alt text.
I do support computer learning to solve issues in a pinch and provide a
bridge or fix for inaccessible content. However, I am hesitant to rely
off of these cloud based solutions as the on demand approach for permanent
accessibility. Cloud based services aren?t reachable everywhere nor in
all situations. While computers may be able to provide you good
information about particular images they may not be able to provide an
equivalent to image without using contextual clues from the text. While
Watson and others may be able to provide this accessibility shouldn?t rely
on paid services that the user doesn?t have control over. A similar
metaphor would be the lifts on buses often break ? instead of retrofitting
accessibility to a bus the bus should be designed to be lower to the
ground to allow access by kneeling rather than require a lift that is
prone to break.
If authoring tools generated accessible content and frameworks had built
in accessibility it would mitigate the effort on authors. I agree there
is more user agent manufacturers need to and can do. However, without
markup it may not always be possible to know the intention of the author.
So creating specifications that allow authors to communicate their
intentions with markup rather than using generic markup is beneficial.
You wouldn?t tell a user without a disability to use a browser plugin to
view a webpage with all the text the same size and color with no spacing
between paragraph in order to make it pleasing to read would you? No,
you?d provide something that is visually formatted for them ? why wouldn?t
we provide something formatted in a way that is adaptable for users with
disabilities as well?
Jonathan
From: Phill Jenkins [mailto:pjenkins@us.ibm.com]
Sent: Thursday, September 22, 2016 4:03 PM
To: w3c-wai-ig@w3.org
Subject: Re: WCAG 2.1
Wayne,
you make a lot of really good points,
some critical needs that may have been overlooked.
My recommendation is to continue on the line of thinking
that "There are user agent problems",
and to get that handful of browsers to not only get those problems fixed,
but,
to address some of the preceived "content author" problems.
Even though there are many problems where
"the solution is entirely within the scope of content authors"
I believe that with some hard work with the browser guys and gals,
that those problems may be even easier solved with better browsers
without external AT.
It just seems to me to be an even more
achievable battle or war
to fight and join the browser guys
than spread ourselves too thin with all the conent authors,
there are just too many of them.
This next example doesn't solve some of the problems you mentioned,
but there are some pretty good APIs available to
AT developers, browser developers, and authoring tool developers
that can already send an image to a software service
to get the text alternative on the fly.
In other words, no need to wait for the content author to make the repair
and add the alt text.
The era of cognitive computing is upon us.
___________
Regards,
Phill Jenkins,
Senior Engineer & Business Development Executive
IBM Research - IBM Accessibility
From: Wayne Dick <waynedick@knowbility.org>
To: Jonathan Avila <jon.avila@ssbbartgroup.com>
Cc: "w3c-wai-ig@w3.org" <w3c-wai-ig@w3.org>
Date: 09/22/2016 01:41 PM
Subject: Re: WCAG 2.1
There are a few points to consider.
- The 2.1 Criteria can strengthen requirements but cannot weaken one. Thus
to meet 2.1 you may change the level from AA to A but you cannot move from
AA to AAA. You may also add additional requirements to an existing SC.
Also, failure to meet 2.1 will not negate conformance to 2.0. If you want
2.1 conformance you must do more. This was in the TF''s and WCAG's
charges.
- At least on the LV task forces we are focusing on content. The ability
to restructure the presentation to a linear arrangement of elements is
completely achievable with content. Resizing text to much more than 200%
with word wrapping is entirely an issue of content. Here are some common
barriers that interfere with access.
* Fixed Top Banners
* Tabular structure that causes lines of enlarged text to run off the page
(Seen frequently in Webmail clients)
* RWD pages that prevent text enlargement
* Inline style specifications that disable cascading style sheets
* Use of background images for anything but decoration
* Multi-column format (Even worse when layout tables are used)
* Margins and padding that resize with font size
* Pages that do not allow restyling
* Use of HTML elements like <select> <option>, and input type radio that
do not support style changes.
Content that interferes with single column reconfiguration, element level
customization, control of line length and enlargement with word wrapping
is inaccessible content, and it is unnecessary.
There are user agent problems. The fact that change of font size loses
your place in the file using almost every user agent is not a problem of
the content author. We know the difference.
- The model of external AT works well for some disabilities. It is not too
effective for others. In the case of providing a single column interface
that enables very large print for people with acuity reduction and short
lines with normal print for people with peripheral field loss the solution
is entirely within the scope of content authors.
Wayne
On Wed, Sep 21, 2016 at 4:19 PM, Jonathan Avila <
jon.avila@ssbbartgroup.com> wrote:
Phil, I had no intention of saying that some criteria were not important
to some users. I agree they are all not just usability issues. They are
barriers to users with disabilities. But as you acknowledge some are
items that must be solved by the author while others can be solved through
the browser or AT. The Understanding Conformance document indicates that
levels are set based on whether the Success Criterion is essential (in
other words, if the Success Criterion isn't met, then even assistive
technology can't make content accessible). So to that degree they carry a
greater weight.
So when I used the term important I was thinking more along these lines ?
some SC are essential because they can?t be worked around by AT and thus
having the author address these is of greater need. The term severity or
priority would probably also be objectionable. But knowing which criteria
affect multiple user groups and which ones can?t be worked around are real
world needs that organizations must consider when there is a limited time
to fix issues. Most organizations do have to prioritize issues whether
something is a barrier versus something AT can overcome is a realistic
factor in helping them make these decisions. While using the term
change level would be best?it carries an ambiguity ? but in the email I
was responding to it appeared that there was confusion about what level
was the minimum level of accessibility. Also many organizations are
confused when we speak of Level AA conformance and automatically assume
incorrectly that excludes level A conformance. So we do need to find a
term that is appropriate and understandable to communicate. What term
would you recommend?
Jonathan
Jonathan Avila
Chief Accessibility Officer
SSB BART Group
jon.avila@ssbbartgroup.com
703.637.8957(Office)
Visit us online: Website | Twitter | Facebook | Linkedin | Blog
Check out our Digital Accessibility Webinars!
From: Phill Jenkins [mailto:pjenkins@us.ibm.com]
Sent: Wednesday, September 21, 2016 1:34 PM
To: w3c-wai-ig@w3.org
Subject: RE: WCAG 2.1
I do not support of the use of phrases such as "raising" or "lowering
importance" or making a success criterion "more" or "less important".
There is a misunderstanding out there about the levels (or its just a
misunderstanding in using English to describe things or me in
understanding things). Saying that it is important today or more
important tomorrow to conform with the success criteria is a very
different concept in my mind that saying that level A success criterion is
more important to this end user than a level triple A success criterion is
to that other user. Remember that it is the web content accessibility
guidelines WCAG , not the end user experience guidelines. In my opinion
all the success criterion are important, period. Level A and double A and
triple A are notabout importance, they are about if and when they apply to
the web content (vs the browser or AT) and to whom do they benefit. One
of the questions that the working group asked themselves when assigning a
level to the success criterion was the following: Does it apply:
to all the content all the time?
to all web sites all the time?
for all audiences all the time?
Another questions that was asked was about whether it is best solved by
the user agent (browser and/or AT) or whether it is best solved by or in
the web content by the web designer and developers. You can read a wide
range of interacting issues that the working group consider in
Understanding Levels of Conformance.
Conformance with triple AAA is not better or more accessibility, it is
more responsibility on the author more of the time because its more
applicability. The notion of "higher levels" (not higher importance) of
conformance comes from that concept that meeting triple AAA requires
passing more, all 65 success criteria, while double AA conformance means
passing fewer only 38 success criteria. Level A is the lowest level,
meaning it only requires meeting 25 success criteria. Even the normative
WCAG standards says:
"Conformance This section is normative. . . . Note 2: It is not
recommended that Level AAA conformance be required as a general policy for
entire sites because it is not possible to satisfy all Level AAA Success
Criteria for some content."
Also, because of conflicts I never recommend requiring many of the Level
AAA success criterion all the time on all web content for all users. Here
are 3 examples:
1. Conforming to 1.4.6Contrast (Enhanced) Level AAA creates a very
high contrast site that is distracting and "striking" for some (many?)
users. "The contrast ratio of 7:1 was chosen for level AAA because it
compensated for the loss in contrast sensitivity [by some]...users...
People with more than [20/80 vision] usually use assistive technologies to
access their content (and the assistive technologies usually have contrast
enhancing, as well as magnification capability built into them). The 7:1
level therefore generally provides compensation for [some users] who do
not [have their] assistive technology. . . ". Therefore, in my opinion,
requiring 7:1 contrast level all the time on all content is akin to
requiring a one size fits all contrast setting that is not in harmony with
the principles of web accessibility and a one size fits one through
transformation technologies in the browsers and AT, not provided by the
web site owner.
2. Conforming to 1.4.9Images of Text (No Exception) Level AAA:
would also require no logos, no images of text, no exceptions. I'm sure
SVG is making progress and being adopted in many places, but I do not
think we are at a point yet that we can require 1.4.9 on all web sites all
the time, hence that is at least one reason it is level AAA. From a
cognitive disability perspective, logos help with branding recognition of
which website the user is on. I would never recommend an "East Berlin"
look (if you ever visited East Berlin before the wall came down you would
know what I mean) for all websites. However, it could make sense to
require this level AAA criterion in some limited cases, for a controlled
set of users, such as a set of AT training pages, hence its level AAA. But
I do not believe sighted users with some cognitive disabilities or aging
users would ever benefit. Most of us are thinking, isn't that the role
of the assistive technology, screen readers, etc, for most web sites and
the answer is of course, hence level AAA.
3. Conforming to 2.4.10Section Headings Level AAA: would require
section headings, "This provision is included at Level AAA because it
cannot be applied to all types of content and it may not always be
possible to insert headings. For example, when posting a pre-existing
document to the Web, headings that an author did not include in the
original document cannot be inserted. Or, a long letter would often cover
different topics, but putting headings into a letter would change the
letter. However, if a document can be broken up into sections with
headings, it facilitates both understanding and navigation". So you see,
in my opinion, section headings are very important to some users for
better comprehension, easier understanding, etc. Its level AAA in my
opinion also because of the difficulty of applying it all the time by the
web owner. New simplification and summarization technology is emerging,
but we would not want to necessarily change the original content all the
time in all cases - hence it would remain a level AAA.
So, moving a success criteria from level A to double A is not necessarily
increasing the importance or benefit to all end users any more than moving
it from level AA to single A is not lowering the importance or benefit to
all end users. However, moving a success criteria from double AA to
single A (or adding a success criteria to level A) *is* about increasing
the amount of responsibility and work onto the web content owner and
increasing its applicability by requiring the criterion to be applied to
all (or more) content, for all (or more) web sites and to all (or more)
audiences all (or more) of the time. Remember that it is the web content
accessibility guidelines WCAG , not the end user experience guidelines.
WCAG conformance is about whether the web content conforms to the standard
by passing the testable success criteria. Browsers, AT, and end users
settings are all still part of the equation. There are also
standards for the browser, called the User Agent Accessibility Guidelines
UAAG, that browsers should conform to. See the discussion on essential
components for more background.
Please, lets stop the misunderstanding that WCAG conformance is the end
all and that the levels are about importance.
Essential Components of Web Accessibility:
https://www.w3.org/WAI/intro/components.php
WCAG 2.0: Conformance Requirements
https://www.w3.org/TR/WCAG20/#conformance-reqs
Understanding WCAG: Understanding Levels of Conformance
https://www.w3.org/TR/UNDERSTANDING-WCAG20/conformance.html#uc-levels-head
___________
Regards,
Phill Jenkins,
IBM Research - IBM Accessibility
ibm.com/able
facebook.com/IBMAccessibility
twitter.com/IBMAccess
ageandability.com
From: Jonathan Avila <jon.avila@ssbbartgroup.com>
To: "w3c-wai-ig@w3.org" <w3c-wai-ig@w3.org>
Date: 09/21/2016 08:21 AM
Subject: RE: WCAG 2.0
Ø So you mean in WCAG2.1 some provisions will be raised from AA to AAA
or A to AA?
It is my personal understanding that the importance of success criteria in
WCAG 2.1 cannot be made less important. So an A would not go to AA. WCAG
2.1 by its definition must ensure that if something passes WCAG 2.1 at a
given level it would also pass WCAG 2 at the same level.
What is a possibility is that a success criteria might get more important.
So a level WCAG 2 AAA criteria might become WCAG 2.1 AA or a WCAG 2 level
AA might become WCAG 2.1 level A.
Jonathan
Jonathan Avila
Chief Accessibility Officer
SSB BART Group
jon.avila@ssbbartgroup.com
703.637.8957(Office)
Visit us online: Website| Twitter| Facebook| Linkedin| Blog
Check out our Digital Accessibility Webinars!
From: Balusani, Shirisha [mailto:sirib@uillinois.edu]
Sent: Wednesday, September 21, 2016 8:58 AM
To: Gregg Vanderheiden; w3c-wai-ig@w3.org
Subject: RE: WCAG 2.0
· Level A: For Level A conformance (the minimum level of
conformance), the Web page satisfiesall the Level A Success Criteria, or a
conforming alternate version is provided.
· Level AA: For Level AA conformance, the Web page satisfies all
the Level A and Level AA Success Criteria, or a Level AA conforming
alternate version is provided.
· Level AAA: For Level AAA conformance, the Web page satisfies all
the Level A, Level AA and Level AAA Success Criteria, or a Level AAA
conforming alternate version is provided.
So you mean in WCAG2.1 some provisions will be raised from AA to AAA or
A to AA?
Thanks,
Siri
From: Gregg Vanderheiden [mailto:gregg@raisingthefloor.org]
Sent: Tuesday, September 20, 2016 11:54 PM
To: Balusani, Shirisha <sirib@uillinois.edu>
Subject: Re: WCAG 2.0
You cannot change a normative part of a standard after it has been issued.
There has been some discussion of a new WCAG 2.1 ? and some discussion in
that about raising some of the provisions from AA to A or AAA to AA
but going from AA to AAA would be lowering an SC?s level
gregg
On Sep 20, 2016, at 10:59 PM, Balusani, Shirisha <sirib@uillinois.edu>
wrote:
Hi All,
I?m curious to know if the WCAG 2.0's success level criteria will be
raised from AA to AAA in near future .
Thanks
Siri
Received on Friday, 23 September 2016 13:28:41 UTC