DOCBOOK-APPS: Bug in stylesheets 1.56, I think

[Gregory Leblanc <GLeblanc at cu-portland dot edu>]

I'm not entirely sure where this bug comes from, but it's because I did
something somewhat strange.  Here's an excerpt from a FAQ that I maintain.
I'm not quite sure where to snip for clarity, so I'll just include the
section, a section of output, and links to the rest of the document and
output.
First, here's by DocBook markup

   <qandaentry>

    <question>
     <para>Why are the RAIDtools at <ulink
url="http://people.redhat.com/mingo/raid-patches/">http://people.redhat.com/
mingo/raid-patches/</ulink> labeled <emphasis>dangerous</emphasis>, and if
they're dangerous, should I use them?</para>
    </question>

    <answer>
     <para>The tools are labeled <emphasis>dangerous</emphasis> because the
RAID code isn't part of the <quote>stable</quote> Linux kernel.</para>
     <para>The tools found at the above URL are the latest and greatest.
You <emphasis>should</emphasis> use these tools with the kernel patches from
the same location.</para>
    </answer>

   </qandaentry>

And here's some html that we get from Jade, run through tidy a couple of
times.

          <dd>
            <dl>
              <dt>3.1. <a href="index.html#AEN95">Why are the
              RAIDtools at <a href=
              "http://people.redhat.com/mingo/raid-patches/"
              target="_top">
              http://people.redhat.com/mingo/raid-patches/</a>
              labeled <i class="EMPHASIS">dangerous</i>, and if
              they're dangerous, should I use them?</a></dt>
            </dl>
          </dd>

As you can see, there is an anchor nested inside of an anchor tag.  From
what the validate at http://www.w3c.org/ tells me, and the way that IE and
Netscape render it, this isn't valid HTML.  Later,
	Greg