E-mail List Archives
Thread: Why Doesn't AT do X?
Number of posts in this thread: 3 (In chronological order)
From: Carol E Wheeler
Date: Mon, Apr 16 2012 11:02AM
Subject: Why Doesn't AT do X?
No previous message | Next message →
I have been wondering about AT, specifically screen readers throughout
several threads. Is the problem mostly too small a pool of users for
real competition? Why DO screen readers do so poor a job?
-- cew
*Carol E. Wheeler*
Web Co-ordinator
American Institute for Cancer Research
1759 R Street NW
Washington DC 20009
Direct Dial: 202-600-3001
Tel: 202-328-7744
Fax: 202-328-7226
= EMAIL ADDRESS REMOVED =
http://www.aicr.org
From: Ryan Hemphill
Date: Mon, Apr 16 2012 12:00PM
Subject: Re: Why Doesn't AT do X?
← Previous message | Next message →
That is a very interesting question. I can only speak to my own
experience, but here goes.
Windows handles accessibility by pushing content to the AT/Screen Reader
with a feed originating from a thing called MSAA. It is, to put it
politely, very crude. In Internet Explorer, this delivery system requires
companies like Freedom Scientific (JAWS) to create hacks all over the place
to make sure its product works. I believe that most of these would be
considered hacks by the dev community. While I have never seen the code
itself, I suspect there are hacks built on hacks built on hacks, resulting
in an unstable piece of software. The other thing is the number of
developers. I believe that practically all of them are understaffed at
best. Another issue stems from the fact that information architecture in
these products falls woefully short and it does not appear that much 'new
blood' is entering these companies that could clean up the problems that
exist.
So that is what I think. Bad a11y support in the OS, too many hacks in the
AT, massive developer team issues and a lack of new perspectives brought in
from outside the current group(s).
It could also be a competition problem, I would agree with that - but I'd
say it is more likely to be based on the details above.
Ryan
On Mon, Apr 16, 2012 at 1:02 PM, Carol E Wheeler < = EMAIL ADDRESS REMOVED = > wrote:
> I have been wondering about AT, specifically screen readers throughout
> several threads. Is the problem mostly too small a pool of users for
> real competition? Why DO screen readers do so poor a job?
>
> -- cew
>
> *Carol E. Wheeler*
>
> Web Co-ordinator
> American Institute for Cancer Research
> 1759 R Street NW
> Washington DC 20009
>
> Direct Dial: 202-600-3001
> Tel: 202-328-7744
> Fax: 202-328-7226
>
> = EMAIL ADDRESS REMOVED =
> http://www.aicr.org
>
> > > >
--
Shipping is a Feature...Perhaps the Most Important Feature.
From: Birkir R. Gunnarsson
Date: Mon, Apr 16 2012 12:15PM
Subject: Re: Why Doesn't AT do X?
← Previous message | No next message
Good discussion.
The only update I'd venture to Ryan's post is that Microsoft has
further complicated the picture by effectivly decomissioning MSAA in
favor of UIA (User Interface Automation), their new standard.
I have no doubt it is more comprihensive and better than MSAA .. but
it still creates a lot of SR work to update older standards, it takes
a lot of evangelism, inside and outside of Microsoft, to make sure
developers use accessible best practices with UIA, and I have found
screen reader developers in general, extremely slow and unresponsive
to bug rports. Add to that the fact that an upgrade containing the
screen reader fix is usually priced well above $100, so the majority
of the end users never end up getting that bug fix, if it is even made
in the first place, until years later.
On top of all that, like we discussed last week, many user's may not
even fully realize the power of their screen reader, and do not take
advantage of the functionality that it does offer, which causes even
further delays in the process of getting something fixed as it relates
to the end user experience.
There are minor desired features, such as being able to monitor a cell
in Excel and have its new value announced as soon as it is updated,
that makes screen readers a lot less useful with that piece of
software.
I have often wondered whether we (or someone) should construct a
database of popular applications along withuser reports of what screen
readers are missing in terms of support for that application.
With such information the vendors might have a clearer idea of what
needs to be prioritized in temrs of features to implement, and users
may be better informed about which screen reader would work with what
application.
I will say, as I have stated previously, that I like the openness and
responsiveness of NvDA's bug tracking and issue logging. It is open to
all and most bugs will get develop comments and mailing list
discussions.
Still NVDA is mostly developed by 2 guys whohave to spend part of
their time adn energy trying to secure funding for the development
itself. This obviously does not allow them adequate resources or time
to address many pressing issues that they could be including in NvDA.
I sincerely hope we can strengthen NVDA, as I feel it is the benchmark
for other and more expensive screen readers, but in general thiis a
tremendously difficult problem to solve, and I feel that assistive
technology companies have often gotten a bit of a free ride when it
comes to accessibility issues, and users are quick to put the blame
squarely on the website or software developer, rather than analyzing
the problem and seeing that shortcomings in the screen reader, or AT
app, are really to blame.
On 4/16/12, Ryan Hemphill < = EMAIL ADDRESS REMOVED = > wrote:
> That is a very interesting question. I can only speak to my own
> experience, but here goes.
>
> Windows handles accessibility by pushing content to the AT/Screen Reader
> with a feed originating from a thing called MSAA. It is, to put it
> politely, very crude. In Internet Explorer, this delivery system requires
> companies like Freedom Scientific (JAWS) to create hacks all over the place
> to make sure its product works. I believe that most of these would be
> considered hacks by the dev community. While I have never seen the code
> itself, I suspect there are hacks built on hacks built on hacks, resulting
> in an unstable piece of software. The other thing is the number of
> developers. I believe that practically all of them are understaffed at
> best. Another issue stems from the fact that information architecture in
> these products falls woefully short and it does not appear that much 'new
> blood' is entering these companies that could clean up the problems that
> exist.
>
> So that is what I think. Bad a11y support in the OS, too many hacks in the
> AT, massive developer team issues and a lack of new perspectives brought in
> from outside the current group(s).
>
> It could also be a competition problem, I would agree with that - but I'd
> say it is more likely to be based on the details above.
>
> Ryan
>
> On Mon, Apr 16, 2012 at 1:02 PM, Carol E Wheeler < = EMAIL ADDRESS REMOVED = > wrote:
>
>> I have been wondering about AT, specifically screen readers throughout
>> several threads. Is the problem mostly too small a pool of users for
>> real competition? Why DO screen readers do so poor a job?
>>
>> -- cew
>>
>> *Carol E. Wheeler*
>>
>> Web Co-ordinator
>> American Institute for Cancer Research
>> 1759 R Street NW
>> Washington DC 20009
>>
>> Direct Dial: 202-600-3001
>> Tel: 202-328-7744
>> Fax: 202-328-7226
>>
>> = EMAIL ADDRESS REMOVED =
>> http://www.aicr.org
>>
>> >> >> >>
>
>
>
> --
>
>
>
> Shipping is a Feature...Perhaps the Most Important Feature.
> > > >