W3C home > Mailing lists > Public > public-html-a11y@w3.org > November 2013

Re: UNS: RE: UNS: WCAG considering amending F65 to NOT fail missing ALT text if title or aria-label is present

From: Sailesh Panchang <spanchang02@yahoo.com>
Date: Mon, 25 Nov 2013 19:51:35 +0000
Message-ID: <1385408899.47858.YahooMailBasic@web125002.mail.ne1.yahoo.com>
To: james nurthen <james.nurthen@oracle.com>
Cc: David MacDonald <david100@sympatico.ca>, Steve Faulkner <faulkner.steve@gmail.com>, HTML Accessibility Task Force <public-html-a11y@w3.org>, WCAG WG <w3c-wai-gl@w3.org>, janina@rednote.net, GV@trace.wisc.edu
Hello James,
An image that does not allow any user interaction is static Web content and not a user interface element covered by the text alternative computation logic in the ARIA specs.
Else surely the authors of the specs would have included an example for such an inactive element to clarify the logic.  plain image that  has no associated interactivity is not what one might term as a"Rich Internet Application" ... the subject of the ARIA specs.
Regards,
Sailesh

--------------------------------------------
On Mon, 11/25/13, james nurthen <james.nurthen@oracle.com> wrote:

 Subject: Re: UNS: RE: UNS: WCAG considering amending F65 to NOT fail missing ALT text if title or aria-label is present
 To: "Sailesh Panchang" <spanchang02@yahoo.com>
 Cc: "David MacDonald" <david100@sympatico.ca>, "Steve Faulkner" <faulkner.steve@gmail.com>, "HTML Accessibility Task Force" <public-html-a11y@w3.org>, "WCAG WG" <w3c-wai-gl@w3.org>, janina@rednote.net, GV@trace.wisc.edu
 Date: Monday, November 25, 2013, 1:04 PM
 
 Sailesh,
 When would images which are displayed to the user ever not
 be user interface elements? I'm not sure I understand the
 basis for this question.
 regards,
 James
 
 
 
 On 11/25/2013 9:08 AM, Sailesh Panchang wrote:
 > Hello All,
 > Can anyone explain the technical basis for recognizing
 aria-labelledby or title as  suitable attributes for
 rendering short text alternative for images that are not
 UIE?
 > The accessible name  (and text alternative)
 computation  logic in ARIA specs [1] is meant only for
 user interface elements.
 > And to ensure this and prevent rist of over /
 mis-interpretation, the ARIA specs defined the term
 'accessible name'  in the context of the ARIA specs
 [2]. So elements (including plain images) that are not UIE
 is out of scope of ARIA specs.
 > Aria-labelledby / aria-label applies to UIE only, not
 plain images. So  the accessible name / text
 alternative computation  logic in the ARIA specs is
 inapplicable to elements that are not UIE.
 > - Aria attributes do not help users who do not use AT
 but yet need  text labels to identify images.
 > - When aria-labelledby is used to label an image and
 the image also has a non-empty alt (to ensure code is
 valid), there is a big ristk that the alt is different from
 aria-labelledby referenced text. This will distort how
 different group of users identifies the image.
 > So when one uses ARIA for purposes it is clearly
 not  intended to be used (as per the Intro to ARIA), it
 is a big big dis-service  to accessibility.
 > This impacts accessibility for real users with
 disabilities who depend on text identifiers for images as
 suggested by
 > Ramón Corominas in another response.
 > So I suggest
 > I. there should be no change to F65  as documented
 currently.
 > ii. there should be no ARIA technique promoting the use
 of aria-labelledby or aria-label on non-UIE elements
 including images.
 > 
 > Thanks,
 > Sailesh Panchang
 > Reference:
 > [1] Text alternative computation    at
 > http://www.w3.org/TR/wai-aria/roles#textalternativecomputation
 > [2] Definition of Accessible name:
 > http://www.w3.org/TR/wai-aria/terms#def_accessible_name
 > 
 > 
 > --------------------------------------------
 > On Mon, 11/25/13, Steve Faulkner <faulkner.steve@gmail.com>
 wrote:
 > 
 >   Subject: Re: UNS: RE: UNS: WCAG
 considering amending F65 to NOT fail missing  ALT text
 if title or aria-label is present
 >   To: "David MacDonald" <david100@sympatico.ca>,
 "Steve Faulkner" <faulkner.steve@gmail.com>,
 "HTML Accessibility Task Force" <public-html-a11y@w3.org>,
 "WCAG WG" <w3c-wai-gl@w3.org>,
 kirsten@can-adapt.com
 >   Date: Monday, November 25, 2013, 5:05
 AM
 >     This http://www.w3.org/html/wg/wiki/ChangeProposals/ImgElement20100504
 >   is a useful detailed account of the
 various arguments for
 >   keeping a strict requirement on alt in
 HTML (for
 >   accessibility reasons).
 >         --
 >     Regards
 >     SteveF
 >   HTML
 >   5.1
 >             On
 25 November 2013
 >   08:31, Steve Faulkner <faulkner.steve@gmail.com>
 >   wrote:
 >       Hi Janina,
 >   I accept
 >   there's a technicality here regarding
 HTML
 >         validation that
 >   makes no judgement whatsoever about
 >   accessibility.
 >   Accessibility advocates argued for 5+
 years
 >   in the html wg against the loosening
 of the requirements on
 >   alt in HTML. It was all about
 accessibility.
 >         --
 >     Regards
 >     SteveF
 >   HTML
 >   5.1
 >           
    On 25 November 2013 01:58,
 >   Janina Sajka <janina@rednote.net>
 >   wrote:
 >         I don't believe
 your analysis is correct. These are not
 >   the opposing
 >     viewpoints. They address
 separate concerns. While I
 >   don't claim to
 >     fully understand what the
 HTML-WG means by "layering
 >   violations," or why those
 >     are a concern, I accept there's
 a technicality here
 >   regarding HTML
 >     validation that makes no
 judgement whatsoever about
 >   accessibility.
 >         Perhaps you and
 others may have been perplexed by James
 >   Craig response
 >     to your first posting on this
 topic this past Friday? His
 >   was the first
 >     response to your post, and
 basically says the same as I
 >   understand what
 >     he wrote:
 >         http://lists.w3.org/Archives/Public/public-html-a11y/2013Nov/0053.html
 >         PS: The 2009 WAI
 Guidance document was not a product of the
 >   HTML-A11Y
 >     Task Force as that TF had not
 yet been created. The document
 >   came from a
 >     special TF that was formed to
 address the specific question
 >   of what HTML
 >     should do regarding alternative
 text, short and long. The TF
 >   in which
 >     both you and I participate
 today was formed later in 2009.
 >   The TF that
 >     created the document cited
 disbanded once the document was
 >   accepted by
 >     the several WAI working groups
 and published.
 >         Janina
 >         David MacDonald
 writes:
 >     > I have no desire to open
 an old debate.  But unless
 >   I’ve missed something HTML5 A11y TF
 2009 resolution and a
 >   2013 A11Y bug response seem to be in
 conflict....
 >     >
 >     >  http://www.w3.org/2009/06/Text-Alternatives-in-HTML5.html
 >     >
 >     > allows aria-labelledby as
 secondary...
 >     >
 >     >
 >     >
 >     > A bug against HTML5 seems
 to have the A11Y TF taking
 >   the opposite position. Unless I’ve
 missed something.
 >     >
 >     >  <https://www.w3.org/Bugs/Public/show_bug.cgi?id=6496>
 >   https://www.w3.org/Bugs/Public/show_bug.cgi?id=6496
 >           >
 >     >
 >     >
 >     > I am willing to go back to
 WCAG with either response
 >   ... I just want to know where the task
 force is ... if it is
 >   not important to the TF, I can go back
 with that also.
 >     >
 >     >
 >     >
 >     > If possible I would like
 WCAG and HTML5 to be
 >   consistent with each other.
 >     >
 >     >
 >     >
 >     > Cheers,
 >     >
 >     > David MacDonald
 >     >
 >     >
 >     >
 >     > CanAdapt Solutions Inc.
 >     >
 >     > Tel:  613.235.4902
 >     >
 >     >  <http://ca.linkedin.com/in/davidmacdonald100>
 >   http://ca.linkedin.com/in/davidmacdonald100
 >           >
 >     >  <http://www.can-adapt.com/>
 >   www.Can-Adapt.com
 >     >
 >     >
 >     >
 >     >   Adapting
 the web to all users
 >     >
 >     >       
      Including those with disabilities
 >     >
 >     >
 >     >
 >     > This e-mail originates
 from CanAdapt Solutions Inc. Any
 >   distribution, use or copying of this
 e-mail or the
 >   information it contains by other than
 the intended
 >   recipient(s) is unauthorized. If you
 are not the intended
 >   recipient, please notify me at the
 telephone number shown
 >   above or by return e-mail and delete
 this communication and
 >   any copy immediately. Thank you.
 >           >
 >     >
 >     >
 >     > Le présent courriel a
 été expédié par CanAdapt
 >   Solutions Inc. Toute distribution,
 utilisation ou
 >   reproduction du courriel ou des
 renseignements qui s'y
 >   trouvent par une personne autre que
 son destinataire prévu
 >   est interdite. Si vous avez reçu le
 message par erreur,
 >   veuillez m'en aviser par téléphone
 (au numéro
 >   précité) ou par courriel, puis
 supprimer sans délai la
 >   version originale de la communication
 ainsi que toutes ses
 >   copies. Je vous remercie de votre
 collaboration.
 >           >
 >     >
 >     >
 >     > From: Sailesh Panchang
 [mailto:spanchang02@yahoo.com]
 >     > Sent: November 24, 2013
 10:23 AM
 >     > To: Steve Faulkner
 >     > Cc: HTML Accessibility
 Task Force; WCAG WG; public-comments-wcag20@w3.org;
 >   Gregg Vanderheiden; Janina Sajka
 >     > Subject: Re: UNS: WCAG
 considering amending F65 to NOT
 >   fail missing ALT text if title or
 aria-label is present
 >     >
 >     >
 >     >
 >     > Hello Steve, I'm saying I
 disagree with the use of
 >   ARIA  for plain  images that
 are not user Interface
 >   elementsHello Steve, I'm saying I
 disagree with the use
 >   of ARIA  for plain  images
 that are not user Interface
 >   elements
 >           >
 >     > Sailesh---
 >     >
 >     > Sent from my iPad ...
 Please pardon
 >   "dictapos" and typos ... <grin>
 >     >
 >     >
 >     > On Nov 24, 2013, at 5:15
 AM, Steve Faulkner <faulkner.steve@gmail.com>
 >   wrote:
 >     >
 >     > Hi sailesh,
 >     >
 >     > what are you saying here?
 >     >
 >     > that you disagree with
 making it OK to use aria-label
 >   etc in place of alt on an image?
 >     > http://lists.w3.org/Archives/Public/public-html-a11y/2013Nov/0052.html
 >     >
 >     > if so then we are in
 aggreement
 >     >
 >     >
 >     >
 >     >
 >     > --
 >     >
 >     > Regards
 >     >
 >     > SteveF
 >     >
 >     > HTML 5.1 <http://www.w3.org/html/wg/drafts/html/master/>
 >     >
 >     >
 >     >
 >     > On 24 November 2013 03:08,
 Sailesh Panchang <spanchang02@yahoo.com>
 >   wrote:
 >     >
 >     > Hello Steve,
 >     >
 >     > 1. Some advance the text
 alternative computation logic
 >   in the ARIA specs as the chief
 motivation for attributes
 >   other than the alt for images,
 specifically the
 >   aria-labelledby and title.
 >     > I find it difficult to
 accept that viewpoint for
 >    reasons noted  in my post:
 >     > http://lists.w3.org/Archives/Public/w3c-wai-gl/2013OctDec/0115.html
 >     >
 >     > 2. As one might expect,
 developers rely on automated
 >   validation checkers to validate
 pages  as suggested by
 >   techniques G134, H88 to ensure
 compliance with SC 4.1.1
 >   (A).
 >     > While only a subset of
 validation rules apply for this
 >   SC, most developers will not be able
 to or do not have
 >   bandwidth to do the fine tuning as
 required for this SC and
 >   will simply aim for full validation as
 the intent to the SC
 >   suggests that content which is
 'created according to the
 >   rules defined in the formal grammar
 for that technology'
 >   is a good thing to ensure
 interoperability and robust
 >   browser/AT support.
 >           > So
 now if one says 'disregard validation errors for
 >   absence of alt attribute, confusion
 will be rife.
 >     > Usefulness of the
 validation checkers too will be
 >   questioned.
 >     > Above all, it is not good
 for the WG to say'it is
 >   fine if one introduces certain types
 of validation issues
 >   into the code'.
 >     >
 >     > Thanks and regards,
 >     >
 >     > Sailesh Panchang
 >     >
 >     >
 --------------------------------------------
 >     >
 >     > On Sat, 11/23/13, Steve
 Faulkner <faulkner.steve@gmail.com>
 >   wrote:
 >     >
 >     >  Subject: Re: UNS:
 WCAG considering amending F65 to
 >   NOT fail missing ALT text  if
 title or aria-label is
 >   present
 >     >
 >     >  To: "David
 MacDonald" <david100@sympatico.ca>,
 >   "HTML Accessibility Task Force" <public-html-a11y@w3.org>,
 >   "WCAG WG" <w3c-wai-gl@w3.org>,
 >   public-comments-wcag20@w3.org,
 >   "Gregg Vanderheiden" <gv@trace.wisc.edu>,
 >   kirsten@can-adapt.com
 >           > 
 Date: Saturday, November 23, 2013, 3:39 AM
 >     >
 >     >
 >     >  Hi Janina,
 >     >  Over time and due to
 experience and understanding,
 >   consensus
 >     >  positions change.
 This document is a useful
 >   historical
 >     >  reference, but does
 not represent the current (lack
 >   of)
 >     >  consensus position
 on the issue.
 >     >
 >     >
 >     >
 >     >  --
 >     >
 >     >  Regards
 >     >
 >     >  SteveF
 >     >  HTML
 >     >  5.1
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >  On 22 November 2013
 >     >  23:54, Janina Sajka
 <janina@rednote.net>
 >     >  wrote:
 >     >
 >     >
 >     >  David:
 >     >
 >     >
 >     >
 >     >  As a point of
 information, the wider WAI community
 >   has
 >     >  already expressed
 >     >
 >     >  a view on this. We
 did so back in 2009, after almost
 >   a year
 >     >  of teleconferences
 nd
 >     >
 >     >  email discussions by
 way of presenting a coherent
 >   approach
 >     >  to the
 >     >
 >     >  HTML-WG.
 >     >
 >     >
 >     >
 >     >  The document we
 produced is entitled, "WAI CG
 >   Consensus
 >     >  Resolutions on
 >     >
 >     >  Text alternatives in
 HTML 5," and is available
 >   at:
 >     >
 >     >
 >     >
 >     >  http://www.w3.org/2009/06/Text-Alternatives-in-HTML5.html
 >     >
 >     >
 >     >
 >     >  So, while it's
 always good to revisit old
 >   thinking, it
 >     >  should not be
 >     >
 >     >  forgotten that we've
 already covered this ground,
 >   and
 >     >  that we covered it
 >     >
 >     >  quite extensively.
 >     >
 >     >
 >     >
 >     >  Janina
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >  David MacDonald
 writes:
 >     >
 >     >  > On behalf of
 the WCAG working group, I have an
 >   action
 >     >  item to solicit
 >     >
 >     >  > responses from
 the wider community regarding a
 >   proposed
 >     >  amendment to WCAG
 >     >
 >     >  > failure
 technique F65 regarding missing ALT.
 >   Currently;
 >     >  if an <img>
 element
 >     >
 >     >  > is missing from
 an ALT attribute the page fails
 >   WCAG SC
 >     >  1.1.1 Level A. Some
 >     >
 >     >  > are proposing
 that we allow authors to use the
 >     >  aria-label,
 aria-labelledby,
 >     >
 >     >  > and title
 attributes INSTEAD of ALT.
 >     >
 >     >  >
 >     >
 >     >  > So under the
 amended failure technique NONE of
 >   the
 >     >  following would
 fail
 >     >
 >     >  > WCAG:
 >     >
 >     >  >
 >     >
 >     >  > <img
 src="../images/giraffe.jpg"
 >     >  title="Giraffe
 grazing on tree
 >   branches"/>
 >     >
 >     >  >
 >     >
 >     >  > <img
 src="../images/giraffe.jpg"
 >     >  aria-label="Giraffe
 grazing on tree
 >     >
 >     >  > branches"/>
 >     >
 >     >  >
 >     >
 >     >  > <img
 src="../images/giraffe.jpg"
 >     > 
 aria-labelledby="123"/>
 >     >
 >     >  > <p
 id="123"> Giraffe grazing on
 >   tree
 >     >  branches</p>
 >     >
 >     >  >
 >     >
 >     >  > As you can
 imagine there are strong opinions all
 >   around
 >     >  on this so I
 >     >
 >     >  > suggested we
 get a sense of what other groups
 >   such as
 >     >  the HTML5 A11y TF
 and
 >     >
 >     >  > PF think.
 >     >
 >     >  >
 >     >
 >     >  > Those in favour
 of the change provide the
 >   following
 >     >  rational:
 >     >
 >     >  >
 >     >
 >     >  > --These
 alternatives on the img element work in
 >     >  assistive
 technology
 >     >
 >     >  > --The aria spec
 says these attributes should get
 >   an
 >     >  accessible NAME in
 the
 >     >
 >     >  > API
 >     >
 >     >  > http://www.w3.org/TR/wai-aria/roles#textalternativecomputation
 >     >
 >     >  > --They say it's
 easy to teach beginner
 >   programmers
 >     >  to just always use
 an
 >     >
 >     >  > aria label on
 everything, rather than requiring
 >   a label
 >     >  on form fields and
 >     >
 >     >  > alt on images
 >     >
 >     >  > --They feel as
 a failure F65 is very strong if
 >   fails a
 >     >  page for missing
 ALT,
 >     >
 >     >  > especially if
 other things work, and they would
 >   like to
 >     >  soften it to allow
 >     >
 >     >  > other things
 that work.
 >     >
 >     >  > --html 5 allows
 a <figure><legend>
 >     >  combination instead
 of alt, so they feel
 >     >
 >     >  > WCAG will have
 to change F65 anyway to allow a
 >   figure
 >     >  with a legend, and
 >     >
 >     >  > that helps open
 the door to this discussion
 >     >
 >     >  >
 >     >
 >     >  > Those in favour
 of the status quo (which fails
 >   missing
 >     >  alt text) provide
 the
 >     >
 >     >  > following
 rational:
 >     >
 >     >  >
 >     >
 >     >  > --aria-label,
 labelledby and title, are not
 >   really
 >     >  suitable attributes
 for
 >     >
 >     >  > img alternative
 text because they implies a
 >   label or
 >     >  title, rather than
 an
 >     >
 >     >  > alternate text,
 so it is not a semantic
 >   equivalent
 >     >
 >     >  > --title is not
 well supported
 >     >
 >     >  > --some feel
 that the aria spec is not in any
 >   way
 >     >  suggesting these as
 >     >
 >     >  > replacements to
 ALT.
 >     >
 >     >  > --aria
 instructs authors to use native html
 >   where
 >     >  possible, and they
 could
 >     >
 >     >  > not come up
 with viable use cases of omitting
 >   alt text
 >     >
 >     >  > --there are
 hundreds of millions of dollars
 >   invested in
 >     >  current evaluation
 >     >
 >     >  > tools, and
 methodologies, and this would
 >   represent a
 >     >  major departure
 from
 >     >
 >     >  > one of the most
 basic accessibility convention,
 >   that is
 >     >  almost as old as
 the
 >     >
 >     >  > web and is the
 "rock star" of
 >   accessibility
 >     >
 >     >  > --it could cost
 a lot of money to change
 >   guidance to
 >     >  developers etc...,
 and
 >     >
 >     >  > muddy the
 waters on a very efficient current
 >   evaluation
 >     >  mechanism
 >     >
 >     >  > --when the
 figure/legend is supported by AT we
 >   can
 >     >  amend F65 but that
 is a
 >     >
 >     >  > different issue
 and the semantics of this
 >   construct are
 >     >  OK for text
 >     >
 >     >  > alternatives,
 rather than the
 >   label/labelledby/title
 >     >  options
 >     >
 >     >  > --it may cause
 some confidence problems to WCAG
 >     >  legislation, because
 it
 >     >
 >     >  > represents a
 strong loosening to a fundamental
 >   Success
 >     >  Criteria, an
 >     >
 >     >  > unnecessary
 change that doesn't help the
 >   cause of
 >     >  accessibility, but
 just
 >     >
 >     >  > complicates
 things
 >     >
 >     >  > --ALT is better
 supported and the text appears
 >   when
 >     >  images are turned
 off.
 >     >
 >     >  > --initial
 twitter feedback from the community
 >   is
 >     >  strongly against
 changing
 >     >
 >     >  > this failure
 >     >
 >     >  >
 >     >
 >     >  >
 >     >
 >     >  > There are
 probably other reasons on both sides
 >   which we
 >     >  hope to hear ...
 but
 >     >
 >     >  > these should
 start it off. Please give your
 >   opinions
 >     >  and reasons.
 >     >
 >     >  >
 >     >
 >     >  > Current
 technique here:
 >     >
 >     >  > http://www.w3.org/TR/WCAG-TECHS/F65.html
 >     >
 >     >  > Proposed
 failure here (see test procedure)
 >     >
 >     >  >
 >     >
 >     >  >
 >     >
 >     >  >
 >     >
 >     >  > Cheers,
 >     >
 >     >  > David
 MacDonald
 >     >
 >     >  >
 >     >
 >     >  > CanAdapt
 Solutions Inc.
 >     >
 >     >  > Tel: 
 613.235.4902
 >     >
 >     >  > http://ca.linkedin.com/in/davidmacdonald100
 >     >
 >     >  >
 www.Can-Adapt.com
 >     >
 >     >  >
 >     >
 >     > 
 >   Adapting the web to all users
 >     >
 >     >  >   
          Including those with
 >     >  disabilities
 >     >
 >     >  >
 >     >
 >     >  >
 >     >
 >     >
 >     >
 >     >  --
 >     >
 >     >
 >     >
 >     >  Janina
 Sajka,   Phone:  +1.443.300.2200
 >   <tel:%2B1.443.300.2200>
 >     >
 >     >       
                
   sip:janina@asterisk.rednote.net
 >   <mailto:sip%3Ajanina@asterisk.rednote.net>
 >           >
 >     >       
           Email:  janina@rednote.net
 >     >
 >     >
 >     >
 >     >  Linux Foundation
 Fellow
 >     >
 >     >  Executive Chair,
 Accessibility Workgroup:       http://a11y.org
 >     >
 >     >
 >     >
 >     >  The World Wide Web
 Consortium (W3C), Web
 >   Accessibility
 >     >  Initiative (WAI)
 >     >
 >     >  Chair, 
 Protocols & Formats     http://www.w3.org/wai/pf
 >     >
 >     >       
   Indie UI
 >    http://www.w3.org/WAI/IndieUI/
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >
 >     >
 >         --
 >         Janina
 Sajka,   Phone:  +1.443.300.2200
 >               
              sip:janina@asterisk.rednote.net
 >               
      Email:  janina@rednote.net
 >         Linux Foundation
 Fellow
 >     Executive Chair, Accessibility
 Workgroup:       http://a11y.org
 >         The World Wide
 Web Consortium (W3C), Web Accessibility
 >   Initiative (WAI)
 >     Chair,  Protocols &
 Formats     http://www.w3.org/wai/pf
 >         
    Indie UI         
               http://www.w3.org/WAI/IndieUI/
 >             
    
 
 
Received on Monday, 25 November 2013 20:09:06 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:05:36 UTC