Another anticipated use for scripting is to automate the new HTML5 canvas element. This element provides an automatable
surface for drawing on, whether it’s done by the content creator (games, animations,
etc.) or the reader (drawing or writing surface), which is why I omitted tackling it
with the rest of the semantics and structure elements.
Although a potentially interesting element to use in ebooks, at this time the canvas element remains largely a black hole to assistive
technologies. A summary of the discussions that have been taking place to fix
the accessibility problems as of writing is available on the Paciello Group website.
Fixes for these accessibility issues will undoubtedly come in time, perhaps directly
for the element or perhaps through WAI-ARIA, but it’s too soon to say.
So is the answer to avoid the element completely until the problems are solved? It would be nice if you could, but wouldn’t be realistic to expect of everyone. Using it judiciously would be a better course to steer.
For now including accessible alternatives is about all you can do. If you’re using
the element to draw graphs and charts, you could add a description with the data
using the aria-describedby attribute and the techniques
we outlined while dealing with images. If you’re using the element for games and the
like, consider the issues we detailed at the outset of the section in determining
how much information to give.
With canvas, we really have to wait and see,
unfortunately.