WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Flash replacement - testing needed

for

From: Terence de Giere
Date: Nov 21, 2004 8:48AM

<EMAIL REMOVED> wrote:

Hi guys... I have been experimenting with a technique called sIFR
that uses a dynamic Flash alphabet to replace headings on a web
page..... However, accessibility was questioned yesterday and I
would REALLY like to know the answer once and for all (I'm getting
two different reports. LOL)

http://mikeindustries.com/blog/files/sifr/2.0/

Stephanie Sullivan

--------------

Stephanie--

No screen readers at the moment, however--

Worked fine in a text-only browser. I have preference bar installed with
Mozilla with a "Flash Killer" button. When that was activated the flash
text vanished but was not replaced, that is, the headings went blank.
With the Opera browser with Flash turned off and JavaScript on, I got a
JavaScript error message, but I have the browser to always show errors.
With Flash on but script off and with script on but Flash off the CSS
styled text displayed. With some tool bars that can turn off CSS (in
Opera and Internet Explorer), both the Flash text and the XHTML text
displayed if CSS was disabled or a user style sheet was employed and
Flash and scripting were on. I don't think you should be concerned about
this because very few browsers will be configured with these add-on tools.

Visually the center column seems to be the lead article, but linearized
as text the left column and then the right column, and finally the
center column is rendered. If possible the code order should reflect the
priorities of the story importance. As a result of the order you have an
h3 heading following an h1 heading rather than an h2. The article at
the end of the code starts with an h2 heading, and h1 to h3 gap should
trigger an error with testing software that checks for W3C Priority 2
accessibility problems. Add character encoding to the page as the W3C
validator needs that information to properly check XHTML, which
otherwise seemed OK.

If the screen reader users do not have problems and these other things
are fixed, this looks like a good solution. I doubt many users are using
a Flash Killer as there is no work-around for this - modem users are
likely to bail out of a page which downloads too slowly. I use it if the
download is slow, so I can see if there is anything on the page besides
flash I might want to look at. In the U.S. about half of home users are
still on dial-up modems and large Flash downloads can be a problem.
About 8% use 28.8K modems, so the amount of Flash that is tolerable may
depend on the target audience, that is, business or home. On my somewhat
slow computer, the page (with Flash) takes about 4 seconds to load on
DSL in various browsers, which means modem use will be over the 8-10
second download recommended by ergonomists. The non-Flash download time
is fine. The ideal time for broadband download is about 1-2 seconds
(based on recommendations by Dr. Jakob Nielsen (Nielsen Norman Group)
and Dr. Eric Schaffer (Human Factors, Inc.).

Terence de Giere
<EMAIL REMOVED>