RE: new failure techniques for GL 1.3

Thanks to both of you!

Comments below.

"Good design is accessible design."

Dr. John M. Slatin, Director 
Accessibility Institute
University of Texas at Austin 
FAC 248C 
1 University Station G9600 
Austin, TX 78712 
ph 512-495-4288, fax 512-495-4524 
email jslatin@mail.utexas.edu 
Web http://www.utexas.edu/research/accessibility 



-----Original Message-----
From: public-wcag-teamb-request@w3.org
[mailto:public-wcag-teamb-request@w3.org] On Behalf Of Loretta Guarino
Reid
Sent: Monday, February 06, 2006 8:32 PM
To: Becky Gibson; public-wcag-teamb@w3.org
Subject: RE: new failure techniques for GL 1.3



> 1.3.2: Failure due to providing text alternatives that include words 
> that appear in a text image, but fail to include information
> that is conveyed
> by the color of certain words in the image [1]

There is no description for this failure. I propose:

"This failure occurs when an image contains text and the color of that
text conveys information, but the text alternatives for the image does
not convey that information." 
[js] Good. But change "alternatives" to "alternative"

> 1.3.3:  Failure due to using CSS to create variations in presentation 
> of text that conveys information without also using
> structure.[2]

This looks ok. Does it make sense to provide a procedure that includes
turning off CSS?
[js] Agree about turning off CSS to test. Also, in the title, "...
Without also using structure" doesn't sound quite right. Do we mean
"without also using structural or semantic markup"? (which is OK if this
is an HTML technique, but not for other technologies)

> 1.3.5: There are two existing failure that still need decisions and 
> one that I think can reuse a 1.3.1 failure.
> 1. Failure due to positioning information with CSS so that
> the visual
> reading order or the programmatically determined reading
> order does not
> convey the meaning of the content. [3] It seems to have a
> failure example
> and a good example - perhaps this should be split into a
> failure technique
> and a positive CSS technique?

Is the good example relevant to the (unpopulated) CSS technique "Using
CSS to reveal reading order"?

The sentence at the beginning of the example should probably be cleaned
up and moved to Description.


> 2. I have already proposed removing this one,  Failure due
> to positioning
> information with HTML layout tables so that the visual reading order 
> or the programmatically determined reading order does not
> convey the meaning
> of the content.  See my proposal at [4].

I don't have enough knowledge about the way HTML is used on the web to
know whether this is still an issue. I think this failure is the
equivalent of the WCAG1 requirement that tables linearize. Is that
really not a problem anymore in practice?
[js] It's certainly less of a problem than it used to be, but still a
problem. And there are a lot of legacy sites out there that do this, so
we may want to keep this as a failure to help with migration to WCAG
2.0. Flagging this as a failure may help people recognize that those
legacy layout tables need to go-- as well as making it clearer that
layout tables aren't a good alternative to CSS positioning.

> 3. I believe that this one (which is currently empty), Failure due to
> creating multiple columns of text in a plain text document
> by including
> lines from both columns, separated by a tab character, in
> a single text
> line, can be replaced with the one I create for 1.3.1,
> Failure due to
> using space characters to create multiple columns in
> simple text content.
> [5].

I agree.
[js] Me too.
> 
> 1.3.6 has two failures:
> 1. Failure due to identifying content only by its shape or location. 
> [6]

This looks good.

> 2. Failure due to using "?"(circle mark glyph) and/or "?"(cross mark
> glyph) alone in order to convey information such as
> "available" or "not
> available", "good" or "bad", "yes" or "no" in the
> comparison table which
> shows the characteristics of the products. [7].  Which I
> suggest renaming
> to, "Failure due to using a non-text mark alone in order
> to convey information".

I like the shorter formulation. (The original title can probably be used
as an additional example.)
[js] Agree that the shorter, more general title is better. 

Will it be clear that this is what "using a non-text mark alone" refers
to? The description almost sounds like a special case of providing alt
text for images.

[js] Is there a word we could use instead of "mark" here? (Or maybe the
"mysteriousness" of "mark" is a good thing-- it could encourage people
to read the failure to find out what it's about...)
Is it possible to mark glyphs with alternative text? I guess you could
use the Supplemental Cues technique, but the wording makes it sound like
there is a way to provide alt text.
[js] The supplemental meaning technique that uses <span> and title won't
solve the problem, at least for now, since screen readers don't read the
title attribute on <span>.

The second example seems like a better example for 1.3.4. 

Loretta

Received on Tuesday, 7 February 2006 13:43:30 UTC