Re: New fundamental test book and covers

Hello,

if SVG as cover causes trouble with some readings-systems, file a bug to the 
developers of such reading-systems - problem solved at the right place.

If the design fits, SVG is pretty good to provide alternatives for graphics,
you can add title, description and metadata for all elements in the file.
Even more: Text elements are already text in the file without accessibility 
problems.
And for description and metadata SVG allows or requires to structure such 
information with elements from other namespaces (I think, the epubchecker
has currently a bug here especially for the content of metadata).
This is much better than the historically poorly designed alt of the 
xhtml:img.

I always use an additional metainformation XHTML file within the content to 
repeat all relevant metadata from the OPF file for the audience, because most 
readings-systems do not manage this.

Printed books typically have a bastard title, half title, colophon to provide 
a text alternative to a graphical cover.
If the raster-cover-information is somehow important, authors need to add such 
an additional and structured XHTML alternative.

Another approach can be to put the raster-cover into an xhtml:figure element 
with an xhtml:figcaption including the structured text description of the 
image.
Just a question of styling to get a nice output for the complete audience.


Gregorio Pellegrino - Fondazione LIA:
...
>as the first element of the spine, an HTML file containing ONLY the cover image 
>(in a rasterized format, not SVG because in our tests did not seem well 
>managed by reading systems and screen readers). 
...

Received on Friday, 7 December 2018 12:40:56 UTC