[Scrollkeeper-devel] omf question

[David M. <da...@lu...>]

I am intentionally crossposting this because IMO this needs to be a
joint conversation. Sorry to those (hopefully) few people who are on
both lists.

Greg Ferguson wrote a perl script to convert docbook sgml metadata
into omf format. It is called, logically enough, db2omf.
I will have some questions about your example omf file and how it
relates to ours.

Here is an annoted version of my first test omf file:


<?xml version="1.0" standalone="no"?>

# your file doesn't include the dtd declaration
<!DOCTYPE omf SYSTEM "OMF.dtd">
<omf xmlns="http://metalab.unc.edu/osrt/omf/" created="20010315" agent="db2omf">

  # I have no idea what the created means. :-)
  # I wrote to Ferg and asked but it's late and he's not at work.
  <resource created="20010315">
    <title created="20010315">Linux Assembly HOWTO</title>
    <date created="20010315">20010215</date>
    <creator created="20010315">
      <person created="20010315">
        <firstName created="20010315">
          Konstantin
        </firstName>
        <lastName created="20010315">
          Boldyshev
        </lastName>
        <email created="20010315">
          ko...@li...
        </email>
      </person>
    </creator>

    # You don't have version information in your example.
    # I dont' know if ours is right yet, either. This is a pre-alpha
    # (non) release.
    <versionGroup created="20010315">
      <version created="20010315">
        <id created="20010315">

          # I suspect this is wrong. Too much info. But it could exist
	  # in the original docbook, too. Haven't checked that yet.
	  # This is why I want to get started doing omf early. There
	  # will be wrinkles to work out. We may have to update all
	  # our documents' source.
          0.6c15 Feb 2001konst
        </id>
        <date created="20010315">
          20010215
        </date>
      </version>
    </versionGroup>

    # Note that we do not have categories. We need your advice.
    # Have you established a system of categorization? I see the
    # <subject><category>System|Other</category></subject>, but I am
    # wondering what the available values are. Can there be multiple
    # categories?
    
    # We are pulling keywords from the title of the document.
    # We will have to come up with a better approach.
    <keywords created="20010315">assembly</keywords>
    <keywords created="20010315">howto</keywords>
    <keywords created="20010315">linux</keywords>

    <description created="20010315">
       This is the Linux Assembly HOWTO, version &version;. This
       document describes how to program in assembly language using
       free programming tools, focusing on development for or from
       the Linux Operating System, mostly on IA-32 (i386) platform.
       Included material may or may not be applicable to other
       hardware and/or software platforms. 
    </description>

    # Are you working from the list established in the latest omf
    # specification? If so, good. I also just asked for a new value,
    # "Quick Reference" since the LDP is going to start working on a
    # series of them. Someday.
    
    <type created="20010315">HOWTO</type>
    
    <format created="20010315" dtd="DocBook" mime="text/sgml" />

    # Ours includes "file://" and yours does not. Which is correct?
    <identifier created="20010315" url="file://home/david/ldp/cvs/LDP/howto/docbook/Assembly-HOWTO.sgml" />

    # Are you using relation, and if so, how? I don't see it in your
    # example.
    <relation created="20010315" url="" />

    # I see Ferg isn't done with this yet, but I did check rfc
    # 1766, and your "C" value should be an ISO language code.
    <language created="20010315" code="EN" name="English" src="" />

  </resource>
</omf>

-- 
Dr. David C. Merrill                     http://www.lupercalia.net
Linux Documentation Project                   da...@lu...
Collection Editor & Coordinator            http://www.linuxdoc.org
                                       Finger me for my public key

"Now this is a totally brain damaged algorithm.  Gag me with a smurfette."
		-- P. Buhr, Computer Science 354