What's the proper way to select a different <imageobject> for HTML and XSL:FO, in a <mediaobject> containing two PNGs? (actually the same PNG but with different scaling)
The way I ended up doing it last year, was by misusing the arch attribute on the imageobjects. An typical <mediaobject> would look like this: <mediaobject> <imageobject arch="html"> <imagedata fileref="images/rh72_press_enter.png" format="PNG"/> </imageobject> <imageobject arch="pdf"> <imagedata fileref="images/rh72_press_enter.png" format="PNG" depth="5cm" scalefit="1"/> </imageobject> </mediaobject> Then I run the attached local-profile.xsl style sheet through Saxon, with the argument "arch=pdf", to create an intermediate DocBook XSL document, which in turn is formatted to XSL:FO using the DocBook XSL style sheets, and finally turned into PDF using Fop. This works, but I don't like this for several reasons: 1. it is misusing the "arch" attribute. However I can't use the more natural "role" for profiling, since I'm using it to get bold for the <emphasis> contents 2. it is not semantic markup. It mixes final processing with the markup 3. it confuses semi-wysiwig editors like xxe and (presumably) Lyx, they show both images, without an indicator that they are alternatives 4. it requires an extra processing steps, with extra intermediate files cluttering up things Thanx! - Steinar