E-mail List Archives
Thread: RE: Recommended method for identifying a line in an "Accessible P DF"
Number of posts in this thread: 2 (In chronological order)
From: julian.rickards
Date: Fri, Feb 21 2003 2:42PM
Subject: RE: Recommended method for identifying a line in an "Accessible P DF"
No previous message | Next message →
<snip>
> Really? I think I must have missed some arguments in the
> discussion. Joe
> Clark's (sic) _book_ suggests alternatives like alt="--" for
> decorative
> horizontal lines, which isn't a good idea in my opinion but
> at least has
> some arguments to support it (text-only browsers). But what
> could possibly
> be the justification of alt="horizontal line"? It looks
> foolish on text
> only browsers. It sounds foolish (and potentially confusing)
> in speech.
> It's completely pointless when rendered by a graphic browser
> in no-images
> mode. What _could_ be the justification?
</snip>
Ok, maybe I made a mistake with regard to the author of the suggestion - I
checked the archives but they are a couple of weeks behind. If I have
attributed the suggestion to the wrong person, my appologies.
However, the original message was with regard to inserting alt text in an
accessible PDF, not a HTML document. My thoughts were that perhaps "---" (my
original alt text) may not be the best alt text so I was prodding this group
for another suggestion. Since the line was not a divider between sections,
"Next section" did not seem right. The line is decorative but without an alt
text, the Accessibility Checker plugin in Acrobat flagged it as being
without alt text.
Jules
----
To subscribe, unsubscribe, or view list archives,
visit http://www.webaim.org/discussion/
From: Terence de Giere
Date: Tue, Feb 25 2003 2:32PM
Subject: Re: Recommended method for identifying a line in an "Accessible PDF"
← Previous message | No next message
Some days ago we were discussing the 'proper' way to identify a
horizontal line in an accessible PDF file, in particular a decorative
line. Presumably the goal is to get the accessibility checker in Acrobat
to sign off with no problems listed.
As with other kinds of accessibility checkers such as for HTML, a good
accessible file may nonetheless generate error messages, or list items
in a page that need to be looked at, but which in the end do not need to
be fixed. For example, the absence of the TABINDEX attribute for HTML
form elements is acceptable if the form elements are logically
sequenced, but the checker will always flag it as a 'problem'. Basically
to do this for images and lines in Acrobat PDF files one needs to add
alternate text (and one can use a blank space if no text is needed for a
decorative item). This will clear the problem with the accessibility
checker. However, Adobe does not think that it is always necessary to
get a clear passing grade with the Acrobat accessibility checker:
"Note that the Accessibility Checker might report problems with elements
in a tagged Adobe PDF file that are things you can safely ignore. For
example, it might report that there are images in the file that do not
have alternate text. However, if these images are just the decorative
borders on the page, they would be unnecessary for someone with a vision
impairment, and would not require alternate text. Similarly, the
Accessibility Checker might report that a running header is not part of
the structure tree. Again, you could leave this as is, since you don't
need this information to be vocalized by a screen reader." ---from "How
to Create Accessible PDF files", Adobe Systems Incorporated
Thus we have the option to ignore the advice of the Acrobat checker, but
normally it is more comfortable if one gets a clear pass because then
others who might analyze the file would not have to pick through the
contested items manually to pass off on the test. This also might bring
up legal issues in some contexts. If the file does not pass an
accessibility test, some slimy lawyer might want to take you to court,
and if this is a possibility, then the less ambiguity in the test
result, the better.
Terence de Giere
= EMAIL ADDRESS REMOVED =