Hi Bob, nice to know someone else is interested in accessibility. What scares me about a new customization layer is the scale of the whole thing. I strongly suspect I'll have some big surprise down the road, so I confess I am thinking about some python postprocessing hacking on already created epub files.
In this project we have to make accessible epubs starting from a very disparate base of already made epub (and a lot of them!), only a few coming from a docbook toolchain (and I do already have some python code for massaging epub files). On the other hand, having an accessible xslt of some sort for DB is something that intrigue me. If I'll at the end follow also a customization layer I'll certainly be willing to share the results, whatever they will be. Time will tell. In the meantime, other interested in the subject could contact me off-list to see if there is some joint effort potential? Thanks, __peppo On Thu, Feb 9, 2012 at 6:44 PM, Bob Stayton <b...@sagehill.net> wrote: > Hi Giuseppe, > The DocBook developers are interested in making DocBook output more > accessible. If you manage to create customizations that improve > accessibility and are willing to share them, I could incorporate the changes > into the base stylesheets for future releases. Any changes that appear to > not be backwards compatible I would probably control with a new stylesheet > parameter ($maximize.accessibility or something ). > > Bob Stayton > Sagehill Enterprises > b...@sagehill.net > > > ----- Original Message ----- From: "Giuseppe Bonelli" > <peppo.bone...@gmail.com> > To: "apps docbook" <docbook-apps@lists.oasis-open.org> > Sent: Wednesday, February 08, 2012 9:58 AM > Subject: [docbook-apps] accessibility and xhtml/html5 xsl > > >> Hi, >> I am working on an epub(3) accessibility project and therefore I have >> had a look at the html5 generated by the latest docbook xslt. >> >> Looking at the html5 generated by the epub3 xsl, I see some tagging >> patterns that are not quite accessible, like the use of <span >> class="italic"> or of structure in the titlepages like: >> >> <body> >> <div class="sect1" title="foo" epub:type="subchapter"> >> <div class="titlepage"> >> <div> >> <div> >> <h2 class="title" style="clear: both" id="idp17104"> >> >> Both of these examples are probably going to confuse screen readers >> and other assistive technologies. >> >> If I remember correctly, the h2 used for section1 titles (therefore >> missing the h1 level) derives from how the xhtml stylesheets have been >> built years ago, so I guess it is not trivial to correct this >> behavior. And I think also the tables will need some reworking (like >> the use of scope and headers attributes for td/th). >> >> My question is: what is the best approach to have a better accessible >> and semantically structured html5 output (particularly for epub3 >> output)? >> >> The two approach that come to my mind are: >> * yet another customization layer; >> * a postprocessing phase done externally of the docbook toolchain by >> means of a python script or the like. >> >> Opinions? >> >> Thanks, >> >> __peppo >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: docbook-apps-unsubscr...@lists.oasis-open.org >> For additional commands, e-mail: docbook-apps-h...@lists.oasis-open.org >> >> >> > --------------------------------------------------------------------- To unsubscribe, e-mail: docbook-apps-unsubscr...@lists.oasis-open.org For additional commands, e-mail: docbook-apps-h...@lists.oasis-open.org