Simplifying image markup

[Norman Walsh <ndw@nwalsh.com>]

Hi folks,

A friend was griping about the complexity of mediaobjects earlier today.

  <figure xml:id="ex.pythagorean">
    <title>The Pythagorean Theorem Illustrated</title>
    <mediaobject>
      <imageobject condition="print">
        <imagedata fileref="figs/print/db5d_ref01.pdf"/>
      </imageobject>
    </mediaobject>
  </figure>

I grant that there are a few more layers of wrapper in there than
you'd really like. The wrappers exist for good reasons, but mostly
related to dealing with multiple output formats (CMYK, hi-res color,
low-res color, etc.).

Those uses are less common today than they were a decade ago.

What I assume my friend wants is an <img> tag. Adding a parallel image
structure to DocBook doesn't appeal to me, but it did occur to me that
we could allow imagedata in more places:

  <figure xml:id="ex.pythagorean">
    <title>The Pythagorean Theorem Illustrated</title>
    <imagedata fileref="figs/print/db5d_ref01.pdf"/>
  </figure>

It leaves a slightly odd element name but it's a single tag. I assume
that audiodata and videodata would be equally useful.

Should we consider allowing "bare" imagedata/audiodata/videodata
elements in DocBook? I suppose they'd be both inline and block
elements.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com>      | We live on an island surrounded by
http://www.oasis-open.org/docbook/ | a sea of ignorance. As our island
Chair, DocBook Technical Committee | of knowledge grows, so does the
                                   | shore of our ignorance.--John 
                                   | Archibald Wheeler

Attachment: signature.asc
Description: PGP signature