[nfbcs] Windows Mouse cursor question

Wed Sep 14 19:01:41 UTC 2016

Debee,

You have hit upon one of the things that scares me as well.  I don't think
we have much to worry about regarding browsing the web and doing email.
There will be options if one considers Windows, IOS, Apple MACs, and
Android.  If something happens in any of those areas, we can move to another
platform even if it is inconvenient.  However, much of the software we use
on the job was made more useable by screen reader hooks.  My job has
required a good deal of work using 3270 emulation.  While I can probably
pick and choose a 3270 emulator that can work with any screen reader, the
two emulators that my employer has used work fine with Window-Eyes and JFW
but do not work with NVDA, most likely because of the more complete OSM's
that those screen reader's had.  If Microsoft comes out with a complete
screen reader but one that is based solely on Microsoft approved techniques,
or if the continued evolution of Windows forces such compliance, I worry
about our ability to use software on the job that doesn't follow the rules.
I think our ability to access software on the job is a real mixed bag.  We
have better access in some ways but poorer access in others.  Some of the
accessibility standards are very reasonable, but some of our access can so
easily be broken.  In many ways, screen readers have not changed that much
over the past twenty years in terms of how software interfaces with them
although a lot has changed under the hood, so to speak, to keep up with
changes to Windows and Microsoft Office.  Much of what has been done,
though, has been done to make sure that we could keep up, not to give us
improved or more flexible access.  We have made a good deal of progress in
terms of awareness of accessibility, but twenty years after the Microsoft
Accessibility Summit, I can't restore the standard functions of function
keys on some HP and other computers without sighted help.  For perhaps five
years, we have not been able to access the envelop addressing dialog in Word
or the signature in Outlook and these were known bugs that did not get
addressed until very recently.  I am seeing a lot of effort being made by
large companies to make software more accessible, but I think it is less the
case that people care about how efficient we can use software.  That can
mean a great deal on the job.  It is a complicated landscape and we need to
be thinking about all this as we move into the future.

Best regards,

Steve Jacobson

-----Original Message-----
From: nfbcs [mailto:nfbcs-bounces at nfbnet.org] On Behalf Of Deborah Armstrong
via nfbcs
Sent: Wednesday, September 14, 2016 1:10 PM
To: NFB in Computer Science Mailing List <nfbcs at nfbnet.org>
Cc: Deborah Armstrong <armstrongdeborah at fhda.edu>
Subject: Re: [nfbcs] Windows Mouse cursor question

Thanks for your well-written description of the issues. I am going to save
it because it's the clearest explanation I've read, and re-assures me that
I'm not going crazy.

I've studied object oriented design so I understand what a model is -- I've
explained to people that the OSM is a database that changes constantly of
what the screen reader believes is onscreen at any given time. Of course
developers know that it contains info that's not currently displayed as
well, but I think this minor over-simplification helps people understand
more clearly what it is.

The idea that there will no longer be a middle ground certainly makes sense.
I've noticed that to script Outlook for example, JAWS now depends on
information exposed from various add-ons. I discovered this when an add-on
got corrupted and prevented the entire JAWS script from working, even though
Outlook, itself was still working fine. 

I suppose whether access depends on a particular add-on or a particular
undocumented operating system hook doesn't in the long run matter, but it's
kind of scary to have it all be so tenuous!

--Debee

-----Original Message-----
From: nfbcs [mailto:nfbcs-bounces at nfbnet.org] On Behalf Of Steve Jacobson
via nfbcs
Sent: Wednesday, September 14, 2016 10:59 AM
To: 'NFB in Computer Science Mailing List'
Cc: Steve Jacobson
Subject: Re: [nfbcs] Windows Mouse cursor question

Deborah,

What you describe frustrates many of us.  In the old days, JAWS and
Window-Eyes used sometimes undocumented hooks to build what they called
their off-screen model of what was being presented visually.  This was often
shortened to OSM.  However, Microsoft has been moving toward a more closed
system requiring that information be retrieved using MSAA or UIA, or
sometimes Document Object Models and the like.  While this works more
reliably in some cases and makes the operating system more stable and secure
in general.  The information we used to get from the OSM is not always
presented in a way that lets us use the mouse keys.  There are cases when
text is available but has a chunk of text without the information associated
with each character, or data is not presented at all in a way that can be
tied to a mouse pointer position.  

Among other problems, this eliminates our ability to sometimes make software
that is not accessible at least useable.  Also, it makes it more difficult
for those of us who write software or web pages to truly know how the
results actually appear.

We have raised this issue with Microsoft and we will continue raising it.
Some of this, though, seems to be a product of us becoming more dependent on
what the operating system or particular software exposes to us.  As I
understand it, some of the same issues exist in the Apple world where there
really has not been an OSM in the same way.  

Having said all this, there are also cases where screen readers have not
fully made available the information that is exposed because it involves
coding for something that is new, and they are challenged to just keep up
with all of the changes.  It is therefore hard to know to what degree screen
readers could be doing more.  Also, we've seen cases where certain UIA calls
that should work cause trouble of their own.  More and more, I am afraid
that software is going to be accessible or not accessible with very little
middleground.

Best regards,

Steve Jacobson

-----Original Message-----
From: nfbcs [mailto:nfbcs-bounces at nfbnet.org] On Behalf Of Deborah Armstrong
via nfbcs
Sent: Tuesday, September 13, 2016 6:05 PM
To: nfbcs at nfbnet.org
Cc: Deborah Armstrong <armstrongdeborah at fhda.edu>
Subject: [nfbcs] Windows Mouse cursor question

I've been using screen readers since the 1980s, and Windows 7 is the first
time I've actually struggled with this issue. I realized this list was the
perfect place to ask.

It appears that the so-called "mouse cursor" (what JFW calls the JAWS
cursor) can no longer freely roam the screen. At first I thought this was
indeed just an issue with JFW, but in experimenting with NVDA and with
WindowEyes I see the same behavior.

I can run an application in Windows XP and explore the entire screen or
active window, depending on whether I restricted the cursor, and pretty much
review everything text-based that is there.

But in Windows 7 (and presumably 8 and 10 as well) half the time what I
receive by exploring the screen with a mouse cursor is a jumble. And from
that jumble text is missing that the screen reader just finished speaking.

This happens on all my machines, in areas where there is no insertion point,
or real cursor. I can run the same software and get two different results
between XP and 7, even with NVDA's screen review feature.

The most dramatic example of this is in Outlook, where in XP I can examine a
message's fields, To, From, date, subject, etc. all using the invisible,
JAWS, mouse or review cursor.

Reviewing the same message in the same version of outlook in Windows 7, only
parts of those fields appear to the mouse cursor.

In a window with multiple panes I could usually get to a pane that didn't
receive focus to read information there. Now it's hit or miss; sometimes I
can read the info, sometimes the screen reader voices it automatically but I
never locate it when I review, and sometimes it's easy to review.

As an advanced user, I always made extensive use of the review capabilities
of my screen reader, and I wish I knew what was going on here and why I
apparently can no longer read everything onscreen. I'd really like a
technical explanation of what is happening and what work-arounds people are
finding?

Thoughts?

--Debee
_______________________________________________
nfbcs mailing list
nfbcs at nfbnet.org
https://urldefense.proofpoint.com/v2/url?u=http-3A__nfbnet.org_mailman_listi
nfo_nfbcs-5Fnfbnet.org&d=CwICAg&c=xoYdONxMEGxjdvKj5bOdEOV28uakaJ20R4TjadGGZB
c&r=gcvya4Pqy0A2EsMRyTgo_3n3PIn53GqWAnSNzbIFuBs&m=nyetaffBCmyHGvaaD43iFqIVGK
UbzYm5jo7IJTx9J3o&s=9SD5TEvESd9I_t_YftFiWhqPqOLgH2nQQZlXTrwfNkw&e=
To unsubscribe, change your list options or get your account info for nfbcs:
https://urldefense.proofpoint.com/v2/url?u=http-3A__nfbnet.org_mailman_optio
ns_nfbcs-5Fnfbnet.org_steve.jacobson-2540visi.com&d=CwICAg&c=xoYdONxMEGxjdvK
j5bOdEOV28uakaJ20R4TjadGGZBc&r=gcvya4Pqy0A2EsMRyTgo_3n3PIn53GqWAnSNzbIFuBs&m
=nyetaffBCmyHGvaaD43iFqIVGKUbzYm5jo7IJTx9J3o&s=_wK5-80paKDl0yVqvrGrA-cM12T2d
vM3r4vWcfktSMA&e= 

_______________________________________________
nfbcs mailing list
nfbcs at nfbnet.org
https://urldefense.proofpoint.com/v2/url?u=http-3A__nfbnet.org_mailman_listi
nfo_nfbcs-5Fnfbnet.org&d=CwICAg&c=xoYdONxMEGxjdvKj5bOdEOV28uakaJ20R4TjadGGZB
c&r=gcvya4Pqy0A2EsMRyTgo_3n3PIn53GqWAnSNzbIFuBs&m=nyetaffBCmyHGvaaD43iFqIVGK
UbzYm5jo7IJTx9J3o&s=9SD5TEvESd9I_t_YftFiWhqPqOLgH2nQQZlXTrwfNkw&e= 
To unsubscribe, change your list options or get your account info for nfbcs:
https://urldefense.proofpoint.com/v2/url?u=http-3A__nfbnet.org_mailman_optio
ns_nfbcs-5Fnfbnet.org_armstrongdeborah-2540fhda.edu&d=CwICAg&c=xoYdONxMEGxjd
vKj5bOdEOV28uakaJ20R4TjadGGZBc&r=gcvya4Pqy0A2EsMRyTgo_3n3PIn53GqWAnSNzbIFuBs
&m=nyetaffBCmyHGvaaD43iFqIVGKUbzYm5jo7IJTx9J3o&s=gL7Hxq91LSRXZ_j3CNVaAjsryNP
RGp-T_4T-Sw0TBTY&e= 

_______________________________________________
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/steve.jacobson%40visi.com