WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: [EXTERNAL]Figures and Captions and Alt-text oh my...

for

From: Sandy Feldman
Date: May 8, 2018 3:12PM

I have been following this thread with real interest.

I am working on a little portfolio site for an artist. Usually alt text
is a tough sell to someone in the visual arts, but this guy has a good
friend who is blind and is right into it.

Each image on the "Paintings" page has alt text, and also a visually
hidden description of the painting provided by the artist. I didn't want
all the back and forth of long description pages. He does not want
people who can see the paintings reading the descriptions.

http://ineeda.coffee/arnie/paintings.html

Does this make sense to anyone but me? Is there a more semantic way to
do this? Comments and suggestions very welcome.

Also - this is a work in progress. The "drawings" page is not ready to
go. Other stuff is still being tweaked.

thanks! Sandy



On 2018-05-08 10:39 AM, Tim Harshbarger wrote:
> I think this brings up an important point.
>
> When you are creating accessibility tests, it is important to understand your goal. Are you trying to test against a standard? Are you testing the user experience? Or are you wanting to test something else?
>
> If you are testing against a standard, you want to make sure the tools and process you use actually tells you if something meets the standard. If you are focused on user experience, then you likely want tests that involve actual users using whatever AT they prefer to use.
>
> A developer just installing a screen reader and listening to find out if it sounds right to the developer typically doesn't achieve either of those goals. If anything, the goal in that situation ends up being to ensure the site sounds good to that developer--which ends up likely creating a site that neither meets a standard or provides a good user experience for those people who do rely on a screen reader. Well, I suppose those things might happen, but they would be accidental outcomes rather than intentional outcomes of the testing.
> -----Original Message-----
> From: WebAIM-Forum [mailto: <EMAIL REMOVED> ] On Behalf Of Birkir R. Gunnarsson
> Sent: Tuesday, May 8, 2018 6:55 AM
> To: WebAIM Discussion List < <EMAIL REMOVED> >
> Subject: [EXTERNAL] Re: [WebAIM] Figures and Captions and Alt-text oh my...
>
> This is what makes NvDA a great screen reader to test with, it does
> much less guess work than, say, Jaws.
>
>
>
> On 5/7/18, Steve Green < <EMAIL REMOVED> > wrote:
>> I recently saw a statement that there are about 1.3 billion websites. I
>> doubt if even 1% of those have been designed and tested to achieve a good
>> level of accessibility. Screen readers therefore have to do the best they
>> can with the other 99%. If some developers are relying on a screen reader to
>> do a quick test, that doesn't seem a good enough reason to impair the user
>> experience on the 99% of website where the developers didn't even test at
>> all.
>>
>> Testing with assistive technologies is important, but developers should not
>> be using screen readers to assess the level of standards compliance of their
>> code - there are much more appropriate tools and techniques for that.
>>
>> Steve
>>
>> -----Original Message-----
>> From: WebAIM-Forum < <EMAIL REMOVED> > On Behalf Of glen
>> walker
>> Sent: 07 May 2018 17:00
>> To: WebAIM Discussion List < <EMAIL REMOVED> >
>> Subject: Re: [WebAIM] Figures and Captions and Alt-text oh my...
>>
>> I agree the user experience would be worse, and the screen is intended for
>> the person that needs it and not the developer, but it also gives a
>> developer a false sense of accomplishment if their code sounds good. For
>> example, jaws will try to find a label for an input field even if the
>> developer didn't code one. That's bad for the developer but good for the
>> end user. Of course, a11y testing should be done on a variety of platforms
>> but there are unfortunately many cases of a quick test and people think
>> they're good.
>>
>> Glen
>>
>> On Mon, May 7, 2018 at 3:40 AM, Steve Green
>> < <EMAIL REMOVED> >
>> wrote:
>>
>>> Professional-grade screen readers have always used heuristics to clean
>>> up bad design - it's one of the key things that sets them apart from
>>> more basic screen readers. The user experience would be far worse if
>>> they did nothing more than present the information that has been coded.
>>>
>>> Steve Green
>>> Managing Director
>>> Test Partners Ltd
>>>
>>>
>>> -----Original Message-----
>>> From: WebAIM-Forum < <EMAIL REMOVED> > On Behalf Of
>>> glen walker
>>> Sent: 07 May 2018 00:18
>>> To: WebAIM Discussion List < <EMAIL REMOVED> >
>>> Subject: Re: [WebAIM] Figures and Captions and Alt-text oh my...
>>>
>>> Screen readers should not have to clean up bad design.
>>>
>>> But that sounds kind of harsh. I would give them the benefit of the
>>> doubt and say they were trying to do the right thing but overdid it.
>>> In any event, a screen reader should not have to fix it. It should
>>> stick with what it does best and present the information that has been
>>> coded.
>>>
>>> Glen
>>>
>>> On Sun, May 6, 2018 at 1:39 PM, Jonathan Cohn < <EMAIL REMOVED> >
>>> wrote:
>>>
>>>> Hello,
>>>>
>>>> I was just reading through the story about Alice's Restaurant that
>>>> was in The Boston Globe this week. While an interesting story, it
>>>> was getting very frustrating by the end with VoiceOver on the
>>>> Macintosh reading essentially the same description for each picture four
>>>> times.
>>>>
>>>> It would read once for the start of the figure once for the end of
>>>> the figure, then the alt text and the text below the picture were also
>>>> read.
>>>> except for the alt attribute on the graphic itself all the text was
>>>> exactly the same. So, our wonderful HTML5 standards have caused in
>>>> at least one screen reader for graphical descriptions to be read four
>>>> times.
>>>>
>>>> OK, I can figure out how to develop a figure that would not be as
>>>> verbose though this In fact, if one puts the Macintosh in "Group"
>>>> web navigation it won't be overly redundant.
>>>>
>>>> But is this issue essentially a Browser / screen reader issue or a
>>>> design issue. I.E. if figure name = caption name = alt-text should
>>>> we be requiring our Screen Reader vendors to clean this up, or
>>>> should the underlying HTML generaed by Newspapers only include
>>>> alternative text if they are not using figure /figcaption?
>>>>
>>>> Best wishes,
>>>>
>>>> Jonathan Cohn
>>>>
>>>>
>>>>