On 24/10/2022 17:01, Norm Tovey-Walsh wrote:
I haven't been paying attention recently: is anyone working on
rendering the original DocBook XML with CSS in the browser, and thus
bypassing the conversion stage altogether?
Not that I’m aware of. For a lot of DocBook, it would be pretty easy.
Yes, the general run of text is not a problem.
You’d have to place elements (for example: title, subtitle, author,
etc.) in the right order because CSS can’t reorder them.
It can't place them earlier than seen, but it can do a surprising amount
of twiddling in situ and post hoc: I did this in my recipes to convert
ingredient data in attributes to human-readable text
(https://balisage.net/Proceedings/vol26/html/Flynn01/BalisageVol26-Flynn01.html)
and I was astonished that it actually works :-)
Any rendering that required reordering (footnotes, epigraph
attributions, some synopsis elements, some mediaobjects) wouldn’t
work.
Footnotes, attributions, and some mediaobjects can be handled (eg as
popups) because HTML doesn't need them re-placed. I don't use synopsis
enough to see the requirements.
I don’t think you could make a multi-page version because counters
would reset on each document (so every chapter would be chapter “1”),
though maybe you could fix that with some clever CSS.
Some trivial scripting could strip any container, cat the chapters, and
enshroud in a new container, but I take the point. I don't think Liam
Quin is on this list, but I've seen what he can do with print-oriented
CSS and it's very impressive.
Peter
