You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There is no benefit yet, and Text::Amuse HTML output is not going to be processed by machine probably (there is a .muse source available, after all), but it might improve accessibility and help search engines index HTML output.
Activity
melmothx commentedon Dec 30, 2017
Agreed, not a big deal and totally worth the effort.
melmothx commentedon Dec 31, 2017
Actually, this is a no-go. Validation barfs everywhere.
https://validator.w3.org/
there is no attribute "role"
http://validator.idpf.org/
Error while parsing file 'attribute "role" not allowed here; expected attribute "class", "dir", "lang", "style", "title" or "xml:lang"'.
link2xt commentedon Dec 31, 2017
I think "role" should be attached to "section", "aside", "p" etc., something that actually holds contents, not h1, h2, ...
See https://stackoverflow.com/questions/4170356/what-values-can-the-html5-xhtml-role-attribute-have
Formal specification of HTML5 explicitly permits it: https://w3c.github.io/html/dom.html#wai-aria
melmothx commentedon Dec 31, 2017
Well, the problem is that the HTML in the epub (version 2) is XHTML strict, not HTML5. Matter needs to be digged.
melmothx commentedon Jan 1, 2018
as far as I can see, EPUB would need to migrate to 3.0 (with extra steps for backcompat) and HTML to HTML5.