Frames, single media rendition

["Dave Pawson" <dave.pawson@gmail.com>]

Spec, 9.3 says

quote.
Frames
A frame is a rectangular container where that contains enhanced
content like text boxes, images or objects. Frames are very similar to
regular drawing shapes, but support some features that are not
available for regular drawing shapes, like contours, image maps and
hyperlinks. In particular, a frame allows to have multiple renditions
of an object. That is, a frame may for instance contain an object as
well as an image. In this case, the application may choose the content
that it supports best. If the application supports the object type
contained in the frame, it probably will render the object. If it does
not support the object, it will render the image.
In general, an application must not render more than one of the
content elements contained in a frame. The order of content elements
dictates the document author's preference for rendering, with the
first child being the most preferred. This means that applications
should render the first child element that it supports. A frame must
contain at least one content element. The inclusion of multiple
content elements is optional. Application may preserve the content
elements they don't render, but don't have to. end quote.


The start of the second paragraph bothers me.

It seems to be aimed at an implementor. A low vision user may best be served
by seeing an image and either reading or listening to a longdesc
(svg:desc) description via AT.

For the current work on captions, it seems most inappropriate?
Given the target of implementors, should we seek a revision of this on
accessibility grounds?

regards


-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk