W3C home > Mailing lists > Public > public-html-a11y@w3.org > September 2012

Re: What does "works" mean here? Re: 48-Hour Consensus Call: InstateLongdesc CP Update

From: Charles McCathie Nevile <chaals@yandex-team.ru>
Date: Mon, 24 Sep 2012 22:42:46 +0200
To: "James Craig" <jcraig@apple.com>
Cc: "HTML Accessibility Task Force" <public-html-a11y@w3.org>
Message-ID: <op.wk5qhkyey3oazb@chaals.local>
[some liberal snipping applied]
On Mon, 24 Sep 2012 18:34:59 +0200, James Craig <jcraig@apple.com> wrote:
> On Sep 23, 2012, at 5:42 AM, Charles McCathie Nevile

>> On Sat, 22 Sep 2012 03:21:10 +0200, James Craig <jcraig@apple.com>

>> As I understand this teachnique, (and from my own testing) it only works
>> for people who are browsing with images off, or using a screen reader.
> That's the target audience, yes.

OK. I suspect this is a really significant point of disagreement. I  
believe that is *one part* of a target audience that is significantly  
larger. That is the TL;DR:


> ... but I still believe the negatives are outweighed by the positive  
> that implementation support for iframe is universal.

Universal support in browsers is great. But I believe that visual  
impairment which leads to ignoring images is less common than visual  
impairment that leads to using images but getting substantial benefit from  
accessibility support designed for non-visual users. (More about that  

That is why I conclude the iframe technique solves the wrong problem, or  
only solves part of the problem. It won't work for a significant number of  
the people who need it. In this respect it is like the various image  
replacement techniques that were designed for screenreader users, but  
failed because they effectively assumed that users more or less divide  
into those who see no images and those who can see with no problems.

By comparison, the woeful level of implementation of longdesc actually  
includes extensions that will work for a huge proportion of those who  
actually need it - with your primary target group already being well  
served by the software they are likely to be using (I am assuming that  
NVDA and VoiceOver do not yet share a majority of the screen reader  
market), even before they look for something tailored to their needs.

I suspect reconciling our differences on who we are trying to help is a  
critical step to reaching consensus on what constitutes an acceptable (set  
of) solutions. The rest of this mail is a partial explanation /  
justification of my belief that the problem space is significantly larger  
than screenreader users.

Be warned. From here are statistics and similar half-truths from a  
snakepit of guesswork, estimation, and weeding through masses of data for  
something that prima facie appears both sufficiently reliable and  
sufficiently detailed to provide some basis for a roughly credible  
statement. On the positive side, I present here all data I found that met  
those criteria.

I was looking for further information on proportions of people with visual  
disability who are blind, compared to those who have vision impairment but  
are not blind, as a partial proxy for people who will use a screenreader,  
or otherwise effectively surf with images off. I believe the data support  
my claim. Numbers are rounded, because statistics are rough truth, but  
with some care because massaging them further isn't helpful.

A UN DISTAT publication from 1990 [1] includes the following:

Austria, 1976: visual impairments 8%, total blindness 0.07%, blind in one  
eye 0.5%
"FRG" (West Germany excluding Bavaria): 270k people with visual  
impairment, of which 50k blind

Then I went looking at the Australian Bureau of Statistics. A couple of  
.xls files give statistics on disability prevalence in the general  
population (as estimated/measured in 2009), and on the distribution of  

The prevalence of "total", "severe", and "moderate" limitation are roughly  
equal at around 600k people of a total population of about 20M [2], with  
the prevalence of "mild" limitation being about double any of those.  
Correlating that with the proportion of disabilities that are related to  
the eye[3], at 3.2% of total limitation, 2.5% of severe limitation and  
0.9% of moderate limitation suggests that total visual limitation is less  
common than moderate to severe limitation (lumping together), which are in  
turn less common than mild limitation.

And then I went back to hunting for data on the US. A lot of broken links,  
but NICHCY (federally funded and focused on US children with disability  
and their education) provides a fact sheet on visual impairment which  

Visual impairment is the consequence of a functional loss of vision,  
rather than the eye disorder itself. Eye disorders which can lead to  
visual impairments can include retinal degeneration, albinism, cataracts,  
glaucoma, muscular problems that result in visual disturbances, corneal  
disorders, diabetic retinopathy, congenital disorders, and infection.


The rate at which visual impairments occur in individuals under the age of  
18 is 12.2 per 1,000. Severe visual impairments (legally or totally blind)  
occur at a rate of .06 per 1,000.

And then there is the WHO's first global report on disability. Which  
recommends that we learn more about it, and figure out how to get on the  
same page (are they reading our mail?), but includes the tidbit [5] That  
in Zambia in 2006, a survey of 28k people suggests .5% cannot see, 2.6%  
have a lot of difficulty seeing, and 4.7% have at least some difficulty.

[1] http://unstats.un.org/unsd/publication/seriesy/seriesy_4e.pdf - around  
page 170 or so (very roughly)
table 1.1
table 2.3
[4] http://nichcy.org/disability/specific/visualimpairment/ - middle of  
the page
[5] http://www.who.int/entity/disabilities/world_report/2011/chapter2.pdf  
page 26 of the report (page 8 of the PDF which is a single chapter of the  
overall report)



Charles McCathie Nevile - Consultant (web standards) CTO Office, Yandex
       chaals@yandex-team.ru         Find more at http://yandex.com
Received on Monday, 24 September 2012 23:56:07 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:05:31 UTC