Re: [Issue 317] Color

To The Group -

There are lots of actual examples of color used in such a way as to make
certain parts of web sites TOTALLY inaccessible.  In the example below (from
www.alaskaair.com) starting with "Fields marked" and ending with
Program/Number", it is unclear to a blind user and possibly a color
deficient user which asterisk is red in color.  In this actual case, the
asterisk after First Name is black - the asterisk after Last Name is red.  A
blind or color deficient user who makes an error on the Alaska Airline web
site will not be abler to correct their errors because color is used alone
to describe what to do.  Is this the type of example we're talking about?

Example follows:

Fields marked with a red asterisk * are required.

1) Traveler  
This is a required field and needs to be completed.
First Name* Last Name* 
Mileage Program/Number

My two cents

Doyle

On 4/21/04 7:03 AM, "John M Slatin" <john_slatin@austin.utexas.edu> wrote:

> 
> Joe, I still find forms where required fields are shown in red without
> additional indication.  Even more often, error alerts may return people
> to forms where the fields that have something wrong with them are shown
> in red but not otherwise identified.
> 
> And no, I don't have a URL in front of me that I canpoint to.
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------
> -
> FIGHT BACK AGAINST SPAM!
> Download Spam Inspector, the Award Winning Anti-Spam Filter
> http://mail.giantcompany.com
> 
> 
> 
> -----Original Message-----
> From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On
> Behalf Of Joe Clark
> Sent: Wednesday, April 21, 2004 9:56 AM
> To: WAI-GL
> Subject: Re: [Issue 317] Color
> 
> 
> 
>> If the text is colored... then it by definition can be read by the
>> User Agent or it wouldn't be able to render it red.
> 
> You mean there's HTML under that rendering? An amazing revelation.
> 
>> If the colored information is a line or something in a graphic - then
>> it would need to be described.
> 
> A poor choice of words.
> 
>> But for a person who is colorblind -- they would generally only have
>> standard visual browsers.    And figuring out which item is what color
> based
>> on color codes in the source would be beyond them.
> 
> and that task *almost never comes up*. I am at the extreme high end of
> Web-surfing habits among Working Group members, as adduced elsewhere,
> and 
> I have found precious few sites that, in essence, say "Click the red box
> 
> to cancel. Click the green box to buy." It borders on nonexistence.
> 
> WCAG WG has this annoying tic of pretending it knows the first thing
> about colour deficiency. Nobody seems to want to do the reading. Not
> only is it readily available, much of it for free, but I brought
> photocopies to the Toronto f2f.
> 
> All we have to do is require authors not to use confusable colours in
> confusable ways. You can put red and green right next to each other and
> it won't necessarily mean anything special for colour-deficient viewers.
> And no, this has *nothing* to do with greyscale display and precious
> little to do with another shibboleth, "contrast."
> 
>> Now
>> 1- we COULD require all color blind people to have a plug in that
>> would expose colors to them.
> 
> Isn't this like denoting a language change by writing "The following
> text 
> is in Spanish"?
> 
>> OR
>> 2-  we COULD require authors to have something that would provide a
>> redundant non-color cue for any color encoded information.
> 
> Every site that uses colour-- and few do not-- has "colour-encoded
> information." Besides, our other proposed guidelines would require
> authors to use something apart from just colour to communicate meaning.
> HTML is not incredibly well-suited to this task, but it can be done. I
> provided examples for the Toronto f2f.
> 
>> Anyone see holes in the above
> 
> It's all hole and no donut.
> 
> 
>> - or additional things to keep in mind?
>> - or anything else on this?
> 
> How about "do your research"? If that seems too onerous, you could
> simply 
> trust the contributor who *has* done it.
> 
> Oh, hi, Kynn. I'm sure you'll have something to say about this, or
> rather, 
> about me.

Received on Wednesday, 21 April 2004 14:23:19 UTC