W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 2011

Re: AW: link text 2.4.4

From: Sailesh Panchang <spanchang02@yahoo.com>
Date: Thu, 1 Dec 2011 06:35:55 -0800 (PST)
Message-ID: <1322750155.57659.YahooMailClassic@web111716.mail.gq1.yahoo.com>
To: adam.solomon2@gmail.com, 'Loretta Guarino Reid' <lorettaguarino@google.com>, Kerstin Probiesch <k.probiesch@googlemail.com>
Cc: 'WCAG' <w3c-wai-gl@w3.org>
It is a failure as per F3 as content is NA when CSS is off so only title will not suffice... maybe it passes 2.4.4 because link has title (i.e. name)  as per H91 but not 1.1.1.

--- On Thu, 12/1/11, Kerstin Probiesch <k.probiesch@googlemail.com> wrote:

From: Kerstin Probiesch <k.probiesch@googlemail.com>
Subject: AW: link text 2.4.4
To: adam.solomon2@gmail.com, "'Sailesh Panchang'" <spanchang02@yahoo.com>, "'Loretta Guarino Reid'" <lorettaguarino@google.com>
Cc: "'WCAG'" <w3c-wai-gl@w3.org>
Date: Thursday, December 1, 2011, 5:11 AM

Thanks for this important discussion. Some cents:

According to the Spec the title-Attribute "offers advisory information about the element for which it is set." When using it for providing very central information (target of a link) it might be a misuse of the attribute. One could say: but alt-attribute is not possible. But one could also ask if there are any situations where it is absolutely necessary to use CSS images and could justify this workaround. Thoughts?

When working with own color schemes those images disappears (F3). Some users describing the situation like this: hunting for Easter eggs without even knowing if there are any. In the first step the user has to know that there are CSS image links and in the next and worst case he has to hunt for them with moving the mouse all over the viewport. If the user was successful he sees the target of the link (title-attribute). If he is not successful he probably missed an important information. So the availability is for a mouse user theoretically given but it depends on the luck or trial and error (and the time). When using own color schemes and the keyboard, the user has nearly no chance - as long as not JS is used. It's very closed to the situation of a blind user when there is neither an alt- nor a title-attribute. The more extensive CSS image is used the more critical it is. I've flickred a screenshot here (a German B2B Site): http://is.gd/rZIbMD. On the
 left side with default colors and on the right side with own color scheme. Because of the border the user will have the idea that there must be "something" (title-attributes are given). Now the user has to move the mouse over this part of page to find out, if there is something. Because of the border the user might think that there is nothing below. There are a lot of same problems on this page: different types of green circles for messages (read, unread) as CSS-images, CSS-images for deleting and answering. Of course not all pages are like this and sometimes it is "just" one image (logo), but those users won't know it. 

Because of that and what is written in the Spec I'm quite unsure if the title-attribute is really sufficient for 1.1.1.



> -----Ursprüngliche Nachricht-----
> Von: Adam Solomon [mailto:adam.solomon2@gmail.com]
> Gesendet: Dienstag, 29. November 2011 22:39
> An: 'Sailesh Panchang'; 'Loretta Guarino Reid'
> Cc: 'WCAG'
> Betreff: RE: link text 2.4.4
> Goof point - F3 in the test procedure states:
> "If an image does convey important information, the information is
> provided to assistive technologies and is also available when the CSS
> image is not displayed."
> Sounds like the title attribute for a link would not constitute
> "availability" when the image is not displayed. Yet, as Loretta pointed
> out (if I understood her correctly), a title attribute might satisfy
> 1.1.1 because of the form control clause, which states that:
> "If non-text content is a control or accepts user input, then it has a
> name that describes its purpose."
> In http://www.w3.org/TR/2010/NOTE-WCAG20-TECHS-20101014/H91 we find
> that the name for a link can be its text, or its title attribute. So, I
> think it possible that the title attribute would suffice for 1.1.1 if
> we apply a liberal meaning to the success criterion.
> -----Original Message-----
> From: Sailesh Panchang [mailto:spanchang02@yahoo.com]
> Sent: Tuesday, November 29, 2011 8:18 PM
> To: 'Loretta Guarino Reid'; adam.solomon2@gmail.com
> Cc: 'WCAG'
> Subject: RE: link text 2.4.4
> Does 'linked text'   include CSS rendered images ... is that covered by
> intent  of SC 2.4.4?
> It is also a SC 1.1.1 failure : see F3.
> Sailesh
> --- On Tue, 11/29/11, Adam Solomon <adam.solomon2@gmail.com> wrote:
> From: Adam Solomon <adam.solomon2@gmail.com>
> Subject: RE: link text 2.4.4
> To: "'Loretta Guarino Reid'" <lorettaguarino@google.com>
> Cc: "'WCAG'" <w3c-wai-gl@w3.org>
> Date: Tuesday, November 29, 2011, 12:26 PM
> So really, for 2.4.4, which requires “The purpose of each link can be
> determined from the link text alone or from the link text together with
> its programmatically determined link context”,  the background-image of
> text would be called the “text” for link purpose, and the title would
> be providing the programmatic name for the link. Thus, no violation,
> but nonetheless a bad practice at best.
> From: Loretta Guarino Reid [mailto:lorettaguarino@google.com]
> Sent: Tuesday, November 29, 2011 6:03 PM
> To: adam solomon
> Cc: WCAG
> Subject: Re: link text 2.4.4
> http://www.w3.org/TR/2010/NOTE-WCAG20-TECHS-20101014/H65 is the nearest
> sufficient technique listed for SC 1.1.1. The use of the title
> attribute generally comes with lots of User Agent Notes, since user
> agent support for title is problematic for low vision and motor
> impaired users.
> On Tue, Nov 29, 2011 at 7:37 AM, adam solomon <adam.solomon2@gmail.com>
> wrote:
> So, since according to http://www.w3.org/TR/2010/NOTE-WCAG20-TECHS-
> 20101014/H91 the name for an anchor tag can be in the title attribute,
> my example would not violate wcag.
> With regard to the point Detlev made about custom stylesheets or
> schemes, is this a violation or a best practice?
> Jon: the example I have is actually a link, but your point is well
> taken, and the link-cancel issue is becoming a widely used practice.
> On Tue, Nov 29, 2011 at 5:20 PM, Loretta Guarino Reid
> <lorettaguarino@google.com> wrote:
> This would be covered by 1.1.1:
> If non-text content is a control or accepts user input, then it has a
> name that describes its purpose. (Refer to Guideline 4.1 for additional
> requirements for controls and content that accepts user input.)
> On Tue, Nov 29, 2011 at 4:39 AM, adam solomon <adam.solomon2@gmail.com>
> wrote:
> Hi everyone
> Does 2.4.4 require there to be link text? Or, in cases where there is
> no link text (for instance a background image of text), a descriptive
> title attribute would suffice, since screen readers (at least Jaws that
> I know for sure) will read the title attribute when no text is present?
Received on Thursday, 1 December 2011 14:36:24 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:34:08 UTC