W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2013

Re: changing presentation of links

From: Charles McCathie Nevile <chaals@yandex-team.ru>
Date: Mon, 25 Mar 2013 13:27:08 +0100
To: "Jonathan Avila" <jon.avila@ssbbartgroup.com>, "Gregg Vanderheiden" <gv@trace.wisc.edu>
Cc: "Alastair Campbell" <alastc@gmail.com>, "W3C WAI ig" <w3c-wai-ig@w3.org>
Message-ID: <op.wuh4vikqy3oazb@chaals.local>
On Sat, 23 Mar 2013 04:36:02 +0100, Gregg Vanderheiden <gv@trace.wisc.edu>  
wrote:

>
> On Mar 22, 2013, at 6:07 PM, Jonathan Avila <jon.avila@ssbbartgroup.com>  
> wrote:
>
>> [Greg wrote] .  It has to also be distinguishable before you point at it
>>
>> And by that do you mean there is a differentiation in color which makes  
>> it distinguishable?
>
> GV: Absolutely not.    Color cannot ever be the only difference.   We  
> even have and SC on that.
>
> But Red and Pink are the same color -- but have different lightness and  
> contrast with each other.   So do any light and dark colors.
>
> Color is HUE and should not be confused with 'lightness'.      If the  
> links are substantially different lightness - it does not matter if they  
> are the same or a different color than the surrounding text.

Until you turn on high-contrast (BLACK AND WHITE) mode.

This ends up getting into questions of how many people you halp by  
default, and whom you leae to sort it out for themselves. Underlining is  
pretty amazingly effective. Font-weight less so. Font face is probably  
less so, since I believe people coerce fonts more frequently than they  
coerce font-weight - but I could be wrong on that, and I have no proof  
either way. Different Hue/Saturation/Lightness probably less so again  
although it varies with the differences, the difference in perceptive  
ability of the user, and user contrast settings.

>> G183 seems to say you are good if there is a non-color differentiation  
>> on focus and on hover.  This certainly raises an interesting point –  
>> it’s not the use of any particular color that conveys meaning here but  
>> the difference in color.
>
> GV:  close but not color,     lightness.

>> I would argue the best test for something like this is not grayscale  
>> but purely black and white because it would require evidence of other  
>> visual changes other than just color difference.  A grayscale check  
>> would potentially pass color differentiation but black and white would  
>> not.
>
> GV:  yes this is close.   but be careful,  if you have colorblindness  
> you don't see the same contrast as a black and white version.   But if  
> you do look at the luminosity of the link and surrounding text (and you  
> can use the contrast tools cited for the contrast provision to measure  
> this)  -- and the contrast between the link text and the other text is  
> enough -- that would be a good enough test according to the techniques   
> -- and that technique was deemed sufficient by the working group.

Hmm. I am not convinced the Working Group got the answer right in that  
case. Maybe we should raise a bug and re-examine the question...

cheers

Chaals

> So - yes -- that it the right idea - but stick with color luminosity  
> contrast rather than greyscale (since there are all sorts of different  
> algorithms for making a color image greyscale -- so greyscale is  
> indeterminate.
>
>>
>> Jonathan
>>
>> From: Gregg Vanderheiden [mailto:gv@trace.wisc.edu]
>> Sent: Friday, March 22, 2013 6:55 PM
>> To: Alastair Campbell
>> Cc: W3C WAI ig
>> Subject: Re: changing presentation of links
>>
>> this issue was just raised in public comment.   If you don't see the  
>> reply there -- you will see the reply shortly.
>>
>> PS - G183 says more than that.  It has to also be distinguishable  
>> before you point at it -- the pointing only makes it more  
>> distinguishable.
>>
>> so the failure and the G183 are not in conflict
>>
>>
>> Gregg
>> --------------------------------------------------------
>> Gregg Vanderheiden Ph.D.
>> Director Trace R&D Center
>> Professor Industrial & Systems Engineering
>> and Biomedical Engineering University of Wisconsin-Madison
>> Technical Director - Cloud4all Project - http://Cloud4all.info
>> Co-Director, Raising the Floor - International -  
>> http://Raisingthefloor.org
>> and the Global Public Inclusive Infrastructure Project -   
>> http://GPII.net
>>
>> On Mar 22, 2013, at 12:42 PM, Alastair Campbell <alastc@gmail.com>  
>> wrote:
>>
>>
>> Hi everyone,
>>
>> Sorry to jump on an old thread, but I'm back in the accessibility
>> universe now, and this came up today.
>>
>> Specifically on link colour and distinguishing links from general
>> text, there's a conflict in the success & fail criteria for 1.4.1.
>>
>> The test from G183 [1] specifically says that changing the link on
>> mouseover/focus to include an underline or other mechanism is ok,
>> whereas F73 [2] says that is not sufficient.
>>
>>
>> From the previous discussion, it sounds like G183 should be
>> deprecated, as if we are talking about "people who cannot perceive
>> color differences" then the contrast aspect is irrelevant.
>>
>> -Alastair
>>
>> 1] http://www.w3.org/TR/2012/NOTE-WCAG20-TECHS-20120103/G183
>> 2] http://www.w3.org/TR/2012/NOTE-WCAG20-TECHS-20120103/F73
>>
>


-- 
Charles McCathie Nevile - Consultant (web standards) CTO Office, Yandex
       chaals@yandex-team.ru         Find more at http://yandex.com
Received on Monday, 25 March 2013 12:27:41 GMT

This archive was generated by hypermail 2.3.1 : Monday, 25 March 2013 12:27:41 GMT