[nfbcs] University of Minnesota Site Accessibility

Doug Lee dgl at dlee.org
Mon Jun 22 11:30:48 UTC 2015 

I work for SSB BART Group, a company in the business of accessibility
testing and remediation. We find that employing persons with
disabilities as testers is an essential part of good accessibility
testing, and that although automation is a goal and a real benefit, it
is just not enough on its own.

It is true that different screen reader behaviors, user habits and
preferences (which can come into play if you're not careful when using
people and specific screen readers to test), and even variations in
other software like Windows, browsers, etc., can severely complicate
the consistency of results and the confidence level in a testing
process. Our solution to this involves a very methodically maintained
database of testing procedures and best practices that are based on all
the experiences and knowledge we have amassed over many years. We in
effect strive to reduce the scaling problems to nothing more complex
than the problem of managing an employee base. I think time has proven
this a successful approach, in terms of both result quality and
scalability.

I understand the difficulty of scaling a process involving numerous
people, divergent requirements and configurations, etc. I also
understand that leaving out these people and divergences will
absolutely result in reduced quality of results. It's a complex
problem. We have one solution. I'm sure there are others out there.
But if there's one thing we have concluded during our years of
accessibility testing, it's surely that you just can't cut out the
audience when designing a product; and that goes for accessibility
remediation just as much as it does for software design in other
contexts.

On Mon, Jun 22, 2015 at 06:17:56AM -0400, NFBCS mailing list wrote:
I'm not talking about an ideal world.  When I worked for the Navy, that's how
the Navy did it and does it with employed screen reader users as and when
staff funding are available.  In fact, it was the United States Navy that
originated this procedure to address the shortcomings it kept on encountering
with software only inspections.

On Sun, 21 Jun 2015, Nicole Torcolini wrote:

>Date: Mon, 22 Jun 2015 01:09:44
>From: Nicole Torcolini <ntorcolini at wavecable.com>
>To: 'Jude DaShiell' <jdashiel at panix.com>,
>    'NFB in Computer Science Mailing List' <nfbcs at nfbnet.org>
>Cc: 'Nancy Coffman' <nancy.l.coffman at gmail.com>
>Subject: RE: [nfbcs] University of Minnesota Site Accessibility
>
>You're talking about an ideal world. Unfortunately, we do not live in an
>ideal world. As I said in my previous email, that is not going to scale. In
>a world where money, time, personnel, expertise, and other factors are not
>constraints, it would work, but, in our world, it will not. It is true that
>software is not going to everything, but at least a little is better than
>none. If a lot of effort is put into it, software can actually do quite a
>bit. It works best if the testing software is designed to work with the
>software that was used to make the website, but, even if this is not so,
>testing software can still be useful.
>Your model has a lot of problems. If an entity knows squat about
>accessibility in the first place, how are they going to hire people that
>they know have the necessary expertise? Then, even if testing was done, the
>results are not of any use unless the tester and the person reading the
>results have a common language. Most people don't know anything about
>virtual buffers, formsmode, ARIA, and the other terminology used in
>accessibility. It is also hard for an external tester to make the results
>understandable unless he/she knows how the underlying infrastructure works;
>not all entities explain how their code works to external testers.
>Most testing software, even if it does not catch everything, at least
>explains errors in terms that most people will understand.
>Finally, if there are conflicting results for six or more screen readers, an
>entity may not know how to fix an error. Even if a testing software does not
>give results that will work for every screen reader, the software usually
>gives results that *should* work with every screen reader, given that the
>browser and screen reader follow the rules. No, it is not ideal, but it is
>scalable.
>I am not doing this just to argue. I am doing this because I work in this
>field, and I know about. I am doing this because, even though I like to
>think about how things should be, I am also a realist, and, frankly, it is
>not really helping if ideas are given to entities that need to do testing
>that are not scalable.
>
>Nicole
>
>-----Original Message-----
>From: Jude DaShiell [mailto:jdashiel at panix.com]
>Sent: Sunday, June 21, 2015 8:38 PM
>To: Nicole Torcolini; 'NFB in Computer Science Mailing List'
>Cc: 'Nancy Coffman'
>
>Subject: RE: [nfbcs] University of Minnesota Site Accessibility
>
>No, there isn't any software that does even part of that.  It's the level of
>expertise a screen reader user has and also software very probably cannot
>detect proprietary controls that show no text.  One thing no software
>package can do reliably is detect when only color is being used to
>communicate meaning.  An error comes up on a form most of the form is blue
>but one entry is red as an example.  The software first has to detect that
>an error state got entered then detect and point out which error shows up in
>what color as opposed to the colors on the rest of the form.  Too much
>intelligence is needed for that.  A form design like that is bad for another
>reason any software package that did testing would likely also fail to point
>out.  When you have exception situations, it helps to clear the rest of the
>screen real estate and show each error separately.  That makes it easier to
>locate and correct.
>
>On Sat, 20 Jun 2015, Nicole Torcolini wrote:
>
>>Date: Sat, 20 Jun 2015 19:07:27
>>From: Nicole Torcolini <ntorcolini at wavecable.com>
>>To: 'NFB in Computer Science Mailing List' <nfbcs at nfbnet.org>
>>Cc: 'Jude DaShiell' <jdashiel at panix.com>,
>>    'Nancy Coffman' <nancy.l.coffman at gmail.com>
>>Subject: RE: [nfbcs] University of Minnesota Site Accessibility
>>
>>Is it any wonder that more accessibility testing is not done? I know
>>that accessibility testing is important, but testing of that type is
>>not going to scale. First, although it is sometimes necessary to get
>>the law involved, it is better to first try just talking about it.
>>Second, a more scalable approach would be to have some sort of
>>software do the testing and give results. No, I do not think that
>>there is any software out there yet that does everything that a human
>>would do, but, sometimes, websites are so bad that almost anything would
>help.
>>
>>Nicole
>>
>>-----Original Message-----
>>From: nfbcs [mailto:nfbcs-bounces at nfbnet.org] On Behalf Of Jude
>>DaShiell via nfbcs
>>Sent: Saturday, June 20, 2015 2:09 PM
>>To: Nancy Coffman via nfbcs
>>Cc: Jude DaShiell; Nancy Coffman
>>Subject: Re: [nfbcs] University of Minnesota Site Accessibility
>>
>>A proper web functional accessibility test needs doing and the test
>>needs to be documented first.
>>The test will need at least one accessibility equipped machine the
>>more the merrier if each machine has different accessibility stuff on
>>it one might have jaws; a second might have Window-Eyes, a third might
>>have Magic and a fourth might have nvda.  Once the offending software
>>can be accessed by each machine the next thing that's needed is for an
>>experienced user to write up use plans to be run on each machine.  The
>>same plan has to be run on each machine.  Next, an experienced user of
>>the accessibility equipment on each of those machines will be needed.
>>Next an accessibility specialist who knows how to do web functional
>>accessibility testing will be needed.  They will ask the user to do
>>everything in the use plan and document what they see and what they
>>hear in writing.  Accessibility violations get documented in that
>>process as well when something doesn't work the same way and
>>communicate over the screen reader or magnification devices on the
>>accessibility testing machines.  The documenting of those tests will
>>need time stamps put on each page.  Next, the offending software's
>>manufacturer's vpat statements for each software package will have to
>>be obtained and compared with the documentation generated by the
>>accessibility tester for accuracy.  Inaccurate vpat's along with
>>documentation of accessibility tool's versions and computer
>>specifications so far as computer performance capabilities and types of
>computers will also have to be documented.
>>Once all of this material is put together and properly typed up, the
>>accessibility specialist can tag each violation with its specific
>>statute sitation.  Once that's been done, it's time to put it into a
>>proper complaint form and put that material up on the Department Of
>>Justice's web accessibility complaint form.  Of course, identities of
>>all participants in the work and dates of performance will also have
>>to be documented.  Once that's done some things should start happening in
>Washington D.C.
>>
>>On Wed, 17 Jun 2015, Nancy Coffman via nfbcs wrote:
>>
>>>Date: Wed, 17 Jun 2015 18:32:42
>>>From: Nancy Coffman via nfbcs <nfbcs at nfbnet.org>
>>>To: NFB in Computer Science Mailing List <nfbcs at nfbnet.org>
>>>Cc: Nancy Coffman <nancy.l.coffman at gmail.com>
>>>Subject: Re: [nfbcs] University of Minnesota Site Accessibility
>>>
>>>PeopleSoft has a long history of selling inaccessible software to
>>>colleges
>>and government agencies. How can we expose them and the way they
>>ignore the law, illeglly selling inaccessible software to agencies
>>with a promise that meets section 508?
>>>
>>>Nancy Coffman
>>>Sent from my iPhone
>>>
>>>>On Jun 11, 2015, at 9:02 AM, Jude DaShiell via nfbcs
>>>><nfbcs at nfbnet.org>
>>wrote:
>>>>
>>>>I tried it with lynx on linux and couldn't do it, but in the class
>>>>search
>>section you have a link that is labeled accessible format which may
>>enable you to do your searching with the technology at hand.
>>>>
>>>>>On Tue, 19 May 2015, Kevin Fjelsted via nfbcs wrote:
>>>>>
>>>>>Date: Tue, 19 May 2015 07:59:07
>>>>>From: Kevin Fjelsted via nfbcs <nfbcs at nfbnet.org>
>>>>>To: NFB in Computer Science Mailing List <nfbcs at nfbnet.org>
>>>>>Cc: Kevin Fjelsted <kfjelsted at gmail.com>
>>>>>Subject: [nfbcs] University of Minnesota Site Accessibility
>>>>>Recently the University of Minnesota upgraded there entire online
>>>>>registration
>>system by purchasing an externally developed software package from
>>PeopleSoft.
>>>>>Before this upgrade I was able with 98% accessibility to search for
>>classes perform online registration tasks, make online book  and fee
>>payments, view my unofficial transcript, and a myriad of other tasks.
>>>>>Since the system change I would estimate that the site is less than
>>>>>ten
>>percent accessible.
>>>>>I have a number of options for browsers and screen readers
>>>>>including
>>JAWS with IE, JAWS with Firefox, Safari with Voiceover both on the Mac
>>and IOS devices, Google Chrome with the Chromevox screenreader
>>extension, NVDA with firefox.
>>>>>Some of the tasks require a log on, however some do not.
>>>>>I would be interested in getting some other feedback and opinions
>>>>>on the
>>site. Is there anyone that can try the "class Search" function?
>>>>>This function should permit one to see the classes offered in a
>>particular term by a particular department.
>>>>>To do this, go to
>>>>>http://onestop.umn.edu <http://onestop.umn.edu/>
>>>>>
>>>>>Under the "tools" section there is a link called "Class Search".
>>>>>
>>>>>This link does not require a log on.
>>>>>Once one has clicked on  this option one must select  the institution.
>>>>>I attempted to choose Twin Cities/Rochester in the combo box.
>>>>>Windows
>>crashed miserably especially with IE. OSX Safari/VO worked the best.
>>Chrome/Chromevox went into repeated loops of talking and froze the
>browser.
>>>>>I then chose "summer 2015" for the term in the next combo box with
>>similar results.
>>>>>I then went to the subject field and typed in "CSCI" for computer
>>science. There seems to be some sort of selector hear from which one
>>is actually supposed to pick a subject, however this access appears to
>>be nonexistent.
>>>>>
>>>>>I then went to the "Show Open Classes Only" field and unchecked
>>>>>this
>>box.
>>>>>On IE the entire system crashes as soon as I uncheck the box.
>>>>>Safari was
>>fine. Chrome froze the speech engine.
>>>>>I Then chose Minneapolis for the campus.
>>>>>The rest of the fields I left as default and clicked on search.
>>>>>Once the search screen came up (which I could only achieve on
>>>>>Safari), I
>>found that the table structure for viewing the data was very difficult
>>to navigate.
>>>>>If anyone can get to this screen on another platform I would
>>>>>appreciate
>>feedback on how easy the class information is to view. For instance,
>>can one see the instructor field for a given class?
>>>>>
>>>>>It appears that the class list is broken up into separate tables
>>>>>for
>>each class which means that one cannot continually navigate down a
>>list of classes with navigation controls. How does this work on other
>platforms?
>>>>>Thanks.
>>>>>-Kevin
>>>>>
>>>>>_______________________________________________
>>>>>nfbcs mailing list
>>>>>nfbcs at nfbnet.org
>>>>>http://nfbnet.org/mailman/listinfo/nfbcs_nfbnet.org
>>>>>To unsubscribe, change your list options or get your account info
>>>>>for
>>nfbcs:
>>>>>http://nfbnet.org/mailman/options/nfbcs_nfbnet.org/jdashiel%40panix.
>>>>>com
>>>>
>>>>--
>>>>
>>>>
>>>>_______________________________________________
>>>>nfbcs mailing list
>>>>nfbcs at nfbnet.org
>>>>http://nfbnet.org/mailman/listinfo/nfbcs_nfbnet.org
>>>>To unsubscribe, change your list options or get your account info
>>>>for
>>nfbcs:
>>>>http://nfbnet.org/mailman/options/nfbcs_nfbnet.org/nancy.l.coffman%4
>>>>0
>>>>gmail.com
>>>
>>>_______________________________________________
>>>nfbcs mailing list
>>>nfbcs at nfbnet.org
>>>http://nfbnet.org/mailman/listinfo/nfbcs_nfbnet.org
>>>To unsubscribe, change your list options or get your account info for
>>nfbcs:
>>>http://nfbnet.org/mailman/options/nfbcs_nfbnet.org/jdashiel%40panix.c
>>>o
>>>m
>>>
>>
>>
>
>

-- 


_______________________________________________
nfbcs mailing list
nfbcs at nfbnet.org
http://nfbnet.org/mailman/listinfo/nfbcs_nfbnet.org
To unsubscribe, change your list options or get your account info for nfbcs:
http://nfbnet.org/mailman/options/nfbcs_nfbnet.org/dgl%40dlee.org

-- 
Doug Lee                 dgl at dlee.org                http://www.dlee.org
SSB BART Group           doug.lee at ssbbartgroup.com   http://www.ssbbartgroup.com
Freedom is not the ability to have what we want.  Freedom is merely the
ability to seek it.  To be free defines what we can do, not what we can get.
(03/28/05)

More information about the NFBCS mailing list