Re: advice on alt text for image maps

2014-01-13 0:57, John Foliot wrote:
> You might consider the *prime* objective of alt text to be "helping 
> people who cannot see the image at all"

I think that has always been the idea.


> but what about other objectives, such as people who can see there is a 
> non-text "blob" on their screen, but lack the visual acuity to discern 
> exactly what the blob is?

Well, let's first consider the specific case under discussion: alt text 
for a map used as image map, in a context where the purpose is simply to 
let the user select one area or another (and obtain information related 
to it). To anyone who does not see acutely, it appears to be far easier 
to make the choice on the basis of the names of the areas rather than 
trying to zoom the image, for example.

Secondary objectives should be treated so that they must not disturb the 
prime objective. Much of this discussion has been about things like 
discussing a page, hypothetically, rather than actual normal use. 
Moreover, it has been about situations where the user can make use of 
the image, at least potentially. In such situations, we can ask why he 
is browsing in a mode where images are not displayed and alt texts are 
presented instead. It sounds like he should have images turned on, so to 
say, and just ask the user agent or assistive software present the alt 
text on request. Or maybe the title attribute.

After all, the alt attribute is supposed to act as replacement for the 
image, not its description. This has been the principle for a long time, 
and it has often been difficult to make authors understand it. Now it 
seems that accessibility-aware people are thinking of alt texts as 
descriptive titles. Paradoxically, this happens now that browsers have 
moved towards more seamless implementation of alt attributes: when 
images are not displayed, an <img> element is often implemented as if 
its alt attribute value appeared in the document instead of the <img> 
element.

HTML5 CR currently has a long and complicated description of the alt 
attribute, partly unrealistic, but the basic principles are good. They 
include the following:

"The most general rule to consider when writing alternative text is the 
following: *the intent is that replacing every image with the text of 
its |alt 
<http://www.w3.org/TR/html5/embedded-content-0.html#attr-img-alt>| 
attribute not change the meaning of the page*.

So, in general, alternative text can be written by considering what one 
would have written had one not been able to include the image.

A corollary to this is that the |alt 
<http://www.w3.org/TR/html5/embedded-content-0.html#attr-img-alt>| 
attribute's value should never contain text that could be considered the 
image's /caption 
<http://www.w3.org/TR/html5/tabular-data.html#the-caption-element>/, 
/title 
<http://www.w3.org/TR/html5/document-metadata.html#the-title-element>/, 
or /legend <http://www.w3.org/TR/html5/forms.html#the-legend-element>/. 
It is supposed to contain replacement text that could be used by users 
/instead/ of the image; it is not meant to supplement the image. The 
|title <http://www.w3.org/TR/html5/dom.html#attr-title>| attribute can 
be used for supplemental information.

Another corollary is that the |alt 
<http://www.w3.org/TR/html5/embedded-content-0.html#attr-img-alt>| 
attribute's value should not repeat information that is already provided 
in the prose next to the image."

Reference: 
http://www.w3.org/TR/html5/embedded-content-0.html#general-guidelines


>
>> How would it help to know that there is a map when you cannot see that
>> map?
> Because it is there. You truly seem to be caught up in a very binary "Blind
> versus Can_See_@_20/20" perspective that surprises me Jukka: alt text is not
> just for blind people.

There is a true dichotomy in the way the alt text has always been 
defined: an <img> element is rendered either by presenting the image 
pointed to by the src attribute or by presenting the alt attribute 
value. Browsers (and authors) have confused this by showing the alt 
attribute value as "tooltip", but that was in the past.

Many authors who have little idea of the meaning and effect of the alt 
attributes but some reason for writing them have routinely written 
descriptions like alt="Large yellow bullet", instead of trying to write 
text that performs the same function as the image. I'm afraid examples 
like map alt="Map of Katoomba"  are very bad examples, as they tend to 
enforce the misunderstanding that authors are supposed to write alt 
texts that describe or comment on images.

There are surely situations where alt="Map of Katoomba", e.g. on a page 
that has such a map simply as a content image, like in an article that 
tells about the history of Katoomba. It is unrealistic to expect authors 
to provide a real alternative text - text that would be really 
equivalent to the map should contain a huge amount of details, in 
addition to well-written high-level description. The best we can do (to 
people who do not see the image) is to announce the presence of such an 
image. But when an <img> element is used for user interface that lets 
the user select between two or more areas, either graphically using an 
image or textually using names of areas, then the alt text should only 
contribute to achieving that purpose.

-- 
Yucca, http://www.cs.tut.fi/~jkorpela/

Received on Sunday, 12 January 2014 23:57:30 UTC