RE: proliferation of reference roles in the dpub aria spec.

jcraig@apple.com wrote: 
> 
> > On Oct 9, 2015, at 6:31 AM, Richard Schwerdtfeger <schwer@us.ibm.com>
> wrote:
> >
> > Why would a cognitively impaired user care that [@rel] maps to a
stylesheet?
> 
> :-)
> 
> This is how you associate style sheets to the current HTML page.
> 
> <link rel="stylesheet" href="screen.css">

I kind of like the @rel idea, as it sits closer to the meta-data side of the
house (which is what I think is being asked for from values such as
biblioentry and glossentry - that seems very 'meta' to me), but we will
still have a struggle getting browsers and those AT tools that may be able
to leverage that data to actually do anything useful with it. For example,
James notes that @rel is used with style-sheets like this:

   <link rel="stylesheet" href="screen.css"> (or more properly, <link
rel="stylesheet" href="screen.css" media="screen">)

And browsers can also reformulate HTML differently when we do this:

   <link rel="stylesheet" href="print.css" media="print">

But ask Wayne Dick about swapping out screen CSS files in the browsers:

   <link rel="stylesheet" href="screen-lowviz.css" media="screen">

{{sad trombone}}.

That said however, I think that banging up browser extensions/plugins to
demonstrate future functionality using an advanced @rel value set seems at
least feasible, but (and this is the key BUT), browsers need to play ball
here - they need to allow for switching (or some other action) based on
attribute values (whether @rel, @aria, @media, or @some_future_thing).
Without the means to act on supplemental author-supplied info, it's all just
hand waving.

JF

Received on Friday, 9 October 2015 16:30:33 UTC