W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2005

Issue Summary 1.4

From: Gregg Vanderheiden <gv@trace.wisc.edu>
Date: Wed, 8 Jun 2005 00:17:30 -0500
To: <w3c-wai-gl@w3.org>
Message-Id: <20050608051728.60C4B1CC3E8@m14.spamarrest.com>

SUMMARY

    The issues are all closed if the following are done

1) algorithm is inserted in guidelines


2) L2 SC3 is changed to 
<proposed>"A mechanism is available to turn off background audio that
plays automatically so that the audio does not interfere with
text-reading software that may be in use.</proposed>


We have a general algorithm but I am still working out the details with
Aries Arditi from Lighthouse.  We have presented an algorithm to a vision
science conference and it was well received but Aries and I are not happy
with its performance at the low end of the spectrum (with foreground or
background black - which has a value of 0 so you can get contrast ratios
that are very high even if the background or foreground is very dark grey
for example). 

Hope to have something soon. In the meantime here are the details assuming
the above to be true (and acceptable). 




DETAILS

-----------------------

ISSUES  [344], [454], [489], [592 ]AND [996]

All just ask for an algorithm. 

DISCUSSION:  one is being developed.   Developed by Arditi and Vanderheiden.
Aries Arditi is with Lighthouse. Van is with Uof Wisc. 

RECOMMENDATION:   
Review  algorithm for adoption when it comes out. 


-----------------------------
ISSUE:    1320. 
Support light text on dark background

As someone who has Retinitis Pigmentosa, I prefer yellow/white text on a 
dark background, but many sites do not address this and present black text 
on a white background, more often than not, with no other options. If 
I've read the guideline correctly, then my concerns are a Priority 3 issue, 
but it is my opinion that text contrast should be more of a priority than 
graphics, as many designers do not try to achieve Priority 3. Is Criteria 
1.4 of WGAC 2.0 trying to address this? 

Loretta says:
I wonder if this is supposed to be covered by 1.3, plus UA requirements?

DISCUSSION:    This is really a user agent issue.  We don't want to
recommend that all sites be light and dark by default.     

RECOMMENDATION:  
Close this issue.  

-----------------------------
ISSUE:   1345. 
GL 1.4, SC L1 : conflicts with good usability
Level 1 Success Criterion for Guideline 1.4: Only requiring that the text
can be programmatically determined implies that it is ok to require the user
to invoke some action (via a user agent or other software) to be able to
read the text. Not very usable.

DISCUSSION:   This is all we can do at level 1 without limiting user
presentation at level 1.  Direct access is at level 2

RECOMMENDATION:  
Close issue. 

-----------------------------
ISSUE: 1372.

 GL 1.4 - discourage background images
Description:

Guideline 1.4 In visual presentations, make it easy to distinguish
foreground words and images from the background.
Should a distinction not be made between background colours (as defined in
the html or CSS) and background images? Should foreground text over
background images not be discouraged? Will the test for measuring the
contrast take into account how people with different types of
colour-blindness see?

DISCUSSION:   
1 - yes - we have a stricter criterion for text over images.  But it is at
level 3
2 - Yes - test is structured so that it works across different types of
color deficiency. 

RECOMMENDATION:  
No Change from current text.  
Close issue

-----------------------------
ISSUE: 1373. 
GL 1.4 SC L1 - remove if redundant with GL 1.1?

Level 1 Success Criteria for Guideline 1.4
If this criterion is met automatically whenever guideline 1.1 is met, should
it be mentioned again as a separate guideline?

DISCUSSION:
1.1 only covers images. This is broader than that.  Note is to make it clear
that this is covered by 1.1.   


RECOMMENDATION:  
No change
Close issue

-----------------------------
ISSUE:  1374. 

GL 1.4, L2 SC 2 - not effective for everyone
Description:

My concerns with this are that:
(a) Guidelines like this exclude a certain percentage of users because the
criterion will invariably be based on x% (90-95%) of users' sight. Is this
not contradicting the principle of universal access?
(b) The measurement is software-based and does not take into account how the
hardware will display the contrast. Different users will use different types
of hardware, e.g. handheld devices and older monitors, which will make it
quite impossible to predict how users will see the contrast.

DISCUSSION:

a) criterion is not based on a % of sight
b) yes - hardware will always affect contrast.  If author follows guidelines
though, the user can pick hardware that will reliably give them a set amount
of contrast. 

RECOMMENDATION:  

Close issue

-----------------------------
ISSUE:  1375. 

GL 1.4, SC 3 Point 2 should be Level 1

Description:
Should Point 2 under Level 3 Success Criteria for Guideline 1.4 not be the
only level 1 success criterion for this guideline? 

DISCUSSION:  Does not relate to current wording.  Mistake or wrong number
or...  

RECOMMENDATION:  
Close - OBE. 


-----------------------------
ISSUE:  1376. 
GL 1.4, SC 3 - incorporate note into SC
Description:

Level 3 Success Criteria for Guideline 1.4, Point 2
Would it not be clearer to non-technical readers if the text in the note was
incorporated in the success criterion and the notion of decibels left out?

DISCUSSION:   Db are measurable so more important to keep as it.  

RECOMMENDATION:  Close issue.


-----------------------------
ISSUE: 1377. 

GL 1.4, SC L3 - impractical to measure background noise
Description:
 
Level 3 Success Criteria for Guideline 1.4, Point 2
How can content authors measure the background noise against the audio
content in practice? Will this guideline work in practice?

DISCUSSION:  There are techniques for doing it.  see techniques doc. 

RECOMMENDATION:  Close issue


-----------------------------
ISSUE: 1430. 

Guideline 1.4, informative - also applies to non-native speakers

It might be worth noting that people not very familiar with a foreign
language have problems understanding speech in the presence of background
noise.

DISCUSSION:  Informative and not accessibility. We decided to not use
usability examples. 

RECOMMENDATION:  Close issue 


-----------------------------
ISSUE:  1440. 

Letting reader control contrast more important than making it measurable
Guideline 1.4. Level 2 Success Criteria. Since this is a rather big problem
it is important to make it measurable but it is yet more important to fulfil
the level 1 Success Criteria so the reader can govern the contrast by him-
or herself.

DISCUSSION:   That is a user agent issue. Out of scope.

RECOMMENDATION:  Close issue. 


-----------------------------
ISSUE:  1498. 

1.4 L2 SC3 - suggested rewording

The following proposal resulted from a report on the impact of not setting
baseline and writing success criterion as functional outcomes:

GL 1.4 level 2 success criterion 3: Users can disable background audio
that plays automatically on a page so that it does not interfere with
text reading software they may be using

Impacted: yes

<proposed>"A mechanism is available to turn off background audio that
plays automatically so that the audio does not interfere with
text-reading software that may be in use.</proposed>

DISCUSSION:  Agree			

RECOMMENDATION:   Adopt rewording. 

-----------------------------
ISSUE: 1081. 

Also, who came up with the 20 dB (or 4 X louder) foreground/background audio
figure? Are there any data to back this up as being sufficient?

DISCUSSION:  Multiple people confirm this value.  From audio industry and
from Gallaudet. 

RECOMMENDATION:   Close issue.  




Gregg

------------------------
Gregg C Vanderheiden Ph.D. 
Professor - Depts of Ind. Engr. & BioMed Engr.
Director - Trace R & D Center 
University of Wisconsin-Madison 
<http://trace.wisc.edu/> FAX 608/262-8848  s
For a list of our list discussions http://trace.wisc.edu/lists/
 
Received on Wednesday, 8 June 2005 05:17:37 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 23:39:37 UTC