escape, not skip, tank traps [was: Re: Inaccessibility of older Flash movies]

At 03:23 AM 2002-08-22, Deneb Meketa wrote:
>Screen readers, for their part, would do well to include a feature that
>allows skipping past a section currently being read.  This already
>appears in the form of, for example, next-paragraph key bindings.
>Perhaps a reasonable enhancement would be a
>skip-past-current-Flash-movie key binding.

This is perhaps an example of where the long experience that stands behind
the design of the digital talking book application is wisdom we should mine.

In the digital talking book, there is an 'escape' method that applies to
forms and tables, and some other selected structural-group elements.  This
allows you to get past something from within it.  The title or label of a
block should be interpreted as being within the structure for the purposes
of this escaping, and to make the interpretation easy that should be the
syntax.  This is consistent with the syntax in ISO HTML and the Digital
Talking Book.

  "Escapable" Structures
  http://www.niso.org/standards/resources/Z39-86-2002.html#Escape

There is a long-standing desire to move toward a normalized architecture
where the content provides the structure and the player provides the methods
for structural navigation.

My point here is that the skip-nav or its cousin here the skip-movie link,
is something to do with what we have today; but that the answer for the
future is not a 'skip' method but an 'escape' method that provides the
'skip' capability and more.  Navigation bars are just another sub-case
along with tables and movies.

The browse method of listening to the title of a section, beginning the
section, and then escaping from the section is just much more robust.  As
compared with trusting the title to completely convey the information needed
to decide to exhaustively browse the section or skip it, 'escape' wins hands
down on usability or user friendliness,  This is why the 'browser back' verb
is such a winner on the Web.  You can sample to the depth you care to and go
back and go elsewhere.  Just within a linearly organized topic tree you
don't need to go back to figure where to go next; the default is to go on to
what's next following the escapable structural unit.

The action plan involves

a) coming to terms with the consumer and player interests that this is a
distinguished class of grouping objects in the structured media, and not all
such containers [involves WCAG, UAAG most]  The PDF solution should prove
that we need this.

b) coming to terms with the format designers on how to indicate this
property with regard to elements in the API or syntax [involves PF, ATIA,
various format groups].  "This property may be thought of as membership in
an escapable class.  If that helps.]

c) deployment of conforming [developer and player] software.

d) deployment of conforming content.

This is a functional deficit of long standing.  It takes an organized
engineering effort to make an effective change.

The PF group is the most likely institutional home for action on this plan.
Those interested in making actual progress on structural navigation please
contact me [off the list suggested] to express this interest.  If we can get
critical mass, we will move out.

Al

http://www.w3.org/WAI/PF

>Thought you might find this interesting -- about Macromedia's choice to
>make older Flash content available to screen readers, and how to work
>around that.
>
>Steve
>
>
>
>-----Original Message-----
>From: Deneb Meketa [mailto:dmeketa@macromedia.com]
>Sent: 21 August 2002 09:27 PM
>Subject: RE: Inaccessibility of older Flash movies
>
>
>Hi Steve -
>
>Thanks for your comments.  It's good to know that people are using our
>accessibility support - even if this leads to problems!
>
>The issues you're describing are typical for older Flash content that is
>complex.  We're looking at a retrofitting problem here: the content was
>developed before we had accessibility support, and now it will take some
>work to make the content accessible.  This is a consequence of our
>choice to try and interpret older content for MSAA.  That choice
>actually has great benefit for simpler legacy content - many Flash
>movies just suddenly start working, without any changes.  The downside
>of this choice is that more complex legacy movies just cause a lot of
>babble.
>
>This kind of problem is not limited to Flash.  Any complex website
>designed before accessibility tools were available would be likely to
>exhibit similar problems.  Many modern DHTML pages fall into this
>category.
>
>There are certainly some problems with our accessibility implementation
>- among these are some things that you mention, like dropdown support
>and tabbing out of movies.  The current Flash Player is really just a
>first version of accessibility support, far from perfect, and we're
>working on major improvements for the next release of the player.
>
>But some of the solution to the larger problem you're talking about
>(complex legacy content) must rest with content maintainers and screen
>readers.
>
>Content maintainers, at the least, need to take a quick look and see
>what their biggest problems are.  If those problems revolve around
>excess verbosity, the near-term strategy is to disable screen reading of
>the movie until there is time to put better accessible logic into the
>movie (if that's even practical at all).  You suggest a perfectly
>reasonable workaround for this near-term solution: provide a skip-Flash
>link.  A slightly better technique that we would endorse is
>re-publishing the Flash movie as version 6, and simply turning off
>accessibility support for the whole movie (uncheck "Make Movie
>Accessible" in the Accessibility panel while nothing is selected on the
>Stage).  This makes the screen reader ignore the Flash content entirely.
>
>Screen readers, for their part, would do well to include a feature that
>allows skipping past a section currently being read.  This already
>appears in the form of, for example, next-paragraph key bindings.
>Perhaps a reasonable enhancement would be a
>skip-past-current-Flash-movie key binding.
>
>Also note that while some of the verbosity can come from the initial
>reading of a Flash movie, a lot can subsequently come from animation in
>the movie prodding the screen reader to re-read the page.  Window-Eyes
>has included a key binding to help users deal with this: if they press
>Ctrl+Shift+F (I think - I might be misremembering the specific keys
>there), Flash events will stop being sent, turning the page into a
>static page that doesn't repeatedly start over from the top.
>
>Thanks again for your input.  We hope that our next version of Flash
>will make enough improvements that everyone will be able to make their
>content accessible without undue effort.
>
>Deneb Meketa
>Flash Player engineering.
>
>
> > -----Original Message-----
> > From: Steve Vosloo [mailto:steve@usabilityjunction.com]
> > Sent: Wednesday, August 21, 2002 5:35 AM
> > To: dmeketa@macromedia.com
> > Subject: Inaccessibility of older Flash movies
> >
> >
> > Hi,
> >
> > I've been playing around with Windows-Eyes and Flash
> > 5 movies. Articles on the MM site say that with Flash Player 6 certain
> > elements of older Flash movies are made available to screen readers --
> > and this is a good thing. But I'm afraid it's not.
> >
> > The problem is that because these older movies were not designed with
> > accessibility in mind, this "feature" can be more of a curse than a
> > blessing, trapping the user in a logically inaccessible movie. The
> > movies I'm looking at are quite complex and very interactive and I can
>
> > assure you it creates pure chaos for the blind user stumbling into the
>
> > movie. The worst part was that once I was in I couldn't tab out of the
>
> > animation.
> >
> > So I've recommended to my client to provide a text description of the
> > Flash movie and also a link to skip over it, e.g.
> >
> > Text description of Flash knee anatomy animation | Skip over Flash
> > animation
> >
> > [Movie goes here]
> >
> > I'm afraid MM still have not created a truly accessible Flash
> > environment (e.g. no drop-down lists yet). Until that happens we'll
> > have to provide a text description and a way around.
> >
> > Thanks
> > Steve
> >
> > Steve Vosloo
> > Division Manager
> > Usability Junction
> >
> > Tel:    + 27 (0) 21 409 7961
> > Fax:   + 27 (0) 21 409 7050
> > Cell:   + 27 (0) 83 463 0012
> > Web:  www.usabilityjunction.com

Received on Thursday, 22 August 2002 09:23:43 UTC