I won't get into a debate with you about the
meaning of interpretation or interpreter. those who may have more
credibility with you may wish to comment. it isn't the main point in the
first place.
I have no objection to a screen-reader showing a
sighted person where the virtual cursor is, but I consider that a teaching aid,
not a practical way to wortk with sighted people.
If I'm talking with a sighted person and they tell
me that something is on the right of the screen, I find that interesting but
useless. I use the find command to find whatever it is. Even if a
screen-reader shows the sighted person where the virtual cursor is, that doesn't
mean the sighted person will be able to tell you how to find it. you have
to use the techniques you use when you browse yourself. the find command
is one of the most underused and best commands to do this in situations such as
we are talking about. if a sighted person tells you that add to card
is halfway down the page, the efficient way for a blind person to find it is not
to use that information. The efficient way is to start at the top of the
page and use the find command.
Also, it just occurred to me that Brian has spoken
often of how things increasingly don't look like what the technical structure
is. So, if someone tells me that add to cart is a button, and it isn't, if
I use the b command and don't find it, that's another example of where the find
command is the best way to find whatever it is. If you know the words you
are searching for accurately, it doesn't matter what a sighted person tells you
it is.
If I'm looking for a link to something like local
listings and a sighted person tells me that it is on the left side of the
screen, a third of the way down, I then know it is probably in the large block
of links a blind person sees at the top of the page. But that doesn't help
me find it efficiently. Searching for the words local listings
does.
Gene
toggle quoted messageShow quoted text
----- Original Message -----
Sent: Thursday, February 21, 2019 11:33 AM
Subject: Re: [nvda] Golden Cursor question
Hi
Travis,
I don’t think Jean either
interacts with sighted people and computers, or he doesn’t know what
interpreting means. One of the two.
All the
best
Steve
What are you talking about in saying my comments aren't
precise?
I said exactly what I meant, if that isn't precise enough for you, then go
bother someone else with trivial matters. If you can't get enough context
from my comments to make intelligent responses, then don't respond, simple as
that.
On 2/21/2019 11:44 AM, Gene wrote:
Expanding abbreviations is the problem of the synthesizer,
not the screen-reader. And what do you mean by your other
comments? they are not precise nor do they take context into
account. Do you use all punctuation all the time? Do you want to
hear every punctuation mark announced? I doubt it. Right there,
you are violating your first statement of what a screen-reader should
do.
----- Original Message -----
Sent: Thursday, February 21, 2019 10:38
AM
Subject: Re: [nvda] Golden Cursor
question
It's always been my opinion that it's a screen reader's job to read the
screen, it's my job to interpret what the screen reader tells me is
there. That means, if there's a graphic, the screen reader should tell
me it's there, if there's an icon, the screen reader should tell me it's
there. The screen reader should read exactly what's there, and nothing
else. I can't tell you how many hours of productivity time I've lost
from stupid little things like my screen reader telling me something was on
the screen, when it wasn't actually there. For example, the voiceover
screen reader (and I think NVDA does as well) says volume, regardless of
whether the entire word is there or not. You know, sometimes, vol by
itself does not mean volume. This is of course a minor example of the
screen reader saying something that isn't there, but you get the idea.
If the screen says v o l, then the screen reader should say v o l, not
volume. I understand that most folks do not agree with me on this, but
it irks the hell out of me when screen readers say things that aren't there,
especially when trying to find sorted file lists, One time I had magazine
issues that were named differently from the publisher, depending on what year
it was published. I spent a good 10 minutes once trying to figure out
why volume 2 issue 1 was out of order, and it turns out the reason was because
the screen reader was reading vol as volume, but the publisher didn't spell
out volume, only put the 3 letters vol, which changed the sorting order of
that particular issue, since the rest of them had volume actually spelled
out. Odd, but there it is. That was 10 minutes that didn't need to
be wasted, because if the screen reader had just read what was on the screen,
it would have been immediately apparent what the problem was, and it could
have been easily corrected.
This is the kind of thing I mean when I say I've lost hours of productivity
due to stupid little things that the screen reader read that didn't
exist.
Although I reported it, and it (eventually) got fixed, at one point, when
beta testing voiceover on the mac, the screen reader said the dinosaur
Stegosaurus as Saint Ego Soars. Again, a case of not reading what's on
the screen, and one that should not have even occurred based on pronunciation
rules built in to the screen reader, but these are the kinds of things that
happen when your screen readers try to interpret things for you. I don't
like it, and I think it's a waste of time. It wastes the user's time,
and it wastes the developer's time, because they have to put in all the rules
that create the speaking rules. Just leave it alone, and let the damned
thing read what's present, let me interpret what it all means.
On 2/20/2019 7:44 AM, Gene wrote:
That's two different questions. changing the format
may be a problem at times, when dealing with sighted people as you
say. but the solution isn't to have the screen-reader not do what it
does. the solution is to teach blind people to find what they are
looking for on the page without being reliant on sighted people's
instructions. for example, if the sighted person tells the person that
add to cart is in the middle of the page, a little up from the center, an
efficient way for a blind person to find it is to disregard the
description. Go to the top of the page and use the screen-reader's
find command to search for add to cart or for cart or whatever you want that
is expected to find the thing efficiently if the search can be done
efficiently.
I may find it interesting when a wwell-intentioned
sighted person tells me where he/she sees something on the screen. I
don't use the description to find the item.
----- Original Message -----
Sent: Wednesday, February 20, 2019 6:27
AM
Subject: Re: [nvda] Golden Cursor
question
I disagree. Changing the layout is interpreting of
course it is. It’s saying well here’s the info, in the format we think
is right, not in the visual format.
Whether you like it or not, that is interpreting and becomes
sometimes, a problem when interacting with sighted people.
All the best
Steve
Changing the layout isn't interpreting. Interpreting
is when you do something like describe something in the screen-reader's own
words. But this doesn't change what is read. It is changing the
layout to make reading logical for blind people. There is nothing
wrong with this. it has been done with great success since MSAA was
introduced in the late Nineties for reading web pages
efficiently.
----- Original Message -----
Sent: Wednesday, February 20, 2019 2:57
AM
Subject: Re: [nvda] Golden Cursor
question
Of course they are interpreting. It reads the screen,
but UIA changes the perceived layout of the screen.
All the best
Steve
What you are complaining about isn't screen
interpreting. it is a decision about what information is included and
how you can move through it. You may disagree with such decisions, but
they aren't screen interpreting.
----- Original Message -----
Sent: Tuesday, February 19, 2019 12:30
PM
Subject: Re: [nvda] Golden Cursor
question
Hi,
OK, incorrect in so much that you can move the mouse, but
only by routing it.
Window-Eyes could not only move by clip, but by graphic as
well, this is really powerful.
If you told me to click on the third graphic from the top
left on the screen, only Window-Eyes would allow me to do
this.
Whatever happened to screen reading, as opposed to screen
interpreting? Don’t even get me started.
All the best
Steve
I know that Window-eyes and JAWS have ways of moving the
mouse such as by pixels and in the case of Window-eyes, in some other way
that I don't recall now, sort of by structure. I remembered the term
while reviewing my message. it is by clip, as Window-eyes called
it. But both also allow for moving the mouse around the screen as you
do in NVDA. The difference is that in NVDA, you move the review
position, then route the mouse to where you stop the review navigator.
In JAWS, you could move the mouse to the word click.
In NVDA, you would move the review navigator to the word click, then route
the mouse. the mouse ends up in the same place. I'm simply
saying that your statement that the mouse can't be moved from the keyboard
in NVDA is factually, not a matter of opinion, not correct. I am not
disagreeing that JAWS and
Window-eyes allow for different, more precise
movements. I'm also not arguing that the Golden Cursor add-on is
necessary in NVDA to move the mouse in finer and more varied ways. But
your statement is factually incorrect.
----- Original Message -----
Sent: Tuesday, February 19, 2019 9:56
AM
Subject: Re: [nvda] Golden Cursor
question
Hi Jean,
But again, although you answered my question, it just proves
to me that I couldn’t find what I wanted, because it is not there, that is
moving the mouse using the keyboard. This is there, if you download
Golden Cursor, but I’ve had this in Window-Eyes for years, so I couldn’t
imagine managing without it now. It even exists in JAWS. Even
mouse search doesn’t currently exist in NVDA, so to me, this isn’t screen
reading at its best at the moment. That’s only to me, remember, just
one guy who is an advanced screen reader user, and in that respect, as I’ve
always said, NVDA, isn’t there yet. I wish it would be.
Another example of this problem is the Izotope plug in
installer. NVDA doesn’t see the screen at all, nothing is read, so you
have to OCR it. When I install with JAWS, it sees the screen, and I
can move the mouse to the Next button, but I can’t click it by tabbing nor
object naving to it, because NVDA simply doesn’t see the screen, unless I
OCR it. JAWS sees it out of the box. This is why I still
maintain that video hooking is a necessary evil.
I know now that many programs use UIA now, so it’s less
important, but the ability to manipulate the mouse via the keyboard, is
still much needed when using custom apps.
All the best
Steve
In the following response, I shall give desktop layout
commands. I don't use the laptop layout and don't know those commands
for what we are discussing.
There aren't specifically mouse movement keys such as in
JAWS. Read the review section of the manual or the relevant
parts. 5.5 is a relevant section. I'm not sure if there are any
others. You will see such commands as num[pad 9, move to next line,
numpad 8, read current line, numpad 7 move to and read previous line.
These are review keys and don't affect the application, they review the
screen. I'm talking about what they do in screen review mode.
They have similar functions when in object navigation but they apply to the
object that has focus.
To move the mouse to the review position, use the command
numpad insert numpad slash. To left click the mouse, use numpad
slash. To right click, use numpad Times, which I believe is also the
asterisk. It's immediately to the right of numpad
slash.
If you can't find how to do something in NVDA, it is not
good methodology or procedure to assume that it can't be done. Asking
here may provide information about how to do it or of an add-on that
does.
Gene
----- Original Message -----
Sent: Monday, February 18, 2019 4:47
AM
Subject: Re: [nvda] Golden Cursor
question
Hello Jean,
So what are the mouse movement keys via the keyboard
then? I’m sorry I can’t find them.
Thanks.
All the best
Steve
You can move the mouse with the keyboard now. You
can't move it as precisely. I don't have an opinion about whether the
Golden Cursor features should be incorporated into the source code.
But your implication that the mouse can't be moved without the Golden Cursor
is not correct.
----- Original Message -----
Sent: Monday, February 18, 2019 2:27
AM
Subject: Re: [nvda] Golden Cursor
question
Jean,
I think the whole Golden Cursor thing should be in NVDA to
be honest. The ability to move the mouse using the keyboard has been
in screen readers, since the invention of Windows.
Supernova has it, System Access has it, JAWS has it,
Window-Eyes was best at it, and so on.
All the best
Steve
The search feature should, I think, be in NVDA, not in the
Golden Cursor. This is important funcionality and is too important to
depend on a user downloading an add-on to have it
available.
----- Original Message -----
Sent: Sunday, February 17, 2019 1:58
PM
Subject: [nvda] Golden Cursor
question
Hi,
In my efforts to find out if Golden Cursor is as good as the
mouse with JAWS, I’d say not quite. Let me explain.
I just downloaded it, and there seems to be no way to search
for a string of text within GC and have the mouse land on that text, so you
can just click it, without routing, saving positions, etc.
Could this possibly be added? A Mouse Search in
NVDA? I use Search in JAWS cursor all the time, and it moves the mouse
to where I want it.
Or am I really stupid and missing it?
Someone suggested that GC does more than the JAWS cursor,
but I don’t really see that.
All the best
Steve
--
Computer Room Services
77 Exeter Close
Stevenage
Hertfordshire
SG1 4PW
Tel: +44(0)1438-742286
Mob: +44(0)7956-334938
Fax: +44(0)1438-759589
Email: steve@...
Web: http://www.comproom.co.uk
|
