Wednesday, April 30, 2008

Metadata and Interoperability

Jim Earley has a very thoughtful post on metadata interoperability: http://jims-thoughtspot.blogspot.com/2008/04/metadata-interoperability.html

As part of DocBook v5, we added the ability to include content from other namespaces in the <info> block to support adding Dublin Core directly in your content. The <info> element's purpose is to house metadata that is not intended for display, so it's a really good fit.

Jim's argument, is that the various standards out there (DocBook, DITA, ODF at a minimum) should move to Dublin Core for metadata, and stop re-inventing the wheel. Dublin Core is an internationally accepted standard for metadata, so why not use it directly?!

I whole-heartedly agree. This approach would add more compatibility between standards, and maybe even facilitate better search! Along with that, Dublin Core is extensible, so it shouldn't be too difficult to add additional metadata fields if you need to.

more on DocBook vs. DITA

Teresa Mulvihill has written an article on DocBook vs. DITA at: http://www.dclab.com/dita_docbook.asp.

It's a well thought-out article, but I'd like to make a few clarifications:

As you may know, Jim Earley and I have presented on this topic and our Doc Standards Interoperability Framework at several conferences, and still hope to form an OASIS TC on document standards interoperability.

In the article, Teresa states:

"DocBook is hierarchical by nature, and must be developed to allow for single-sourced content. DocBook has a fixed element and attribute set."

I've successfully used and recommended single-sourcing approaches with DocBook, without additional development. It's quite easy to set up a book or article and use XIncludes or even file entity references to pull in content from a common pool of content structures (usually section or chapter).

It's also fairly easy to extend the elements and attributes in DocBook. This has been made even easier with DocBook v5.0 and RelaxNG. In fact, the DocBook Subcommittee for Publishers that I chair, has helped organize the source patterns for DocBook v5 with a modular approach, enabling easier customizations to be created. Our subcommittee has created a customization geared specifically to publishers, without all of the technical blocks and inlines in full DocBook. This significantly reduces the tag set for folks that do not produce software or technical documentation to use the DocBook standard for general publishing!

DocBook can generate more than PDF, HTML and HTMLHelp. The docbook-xsl-1.73.2 stylesheet distribution supports: html, htmlhelp, javahelp, manpages, xhtml, Word roundtrip, slides and websites!

I would also argue that DocBook can be used on very high volume documentation projects, as well as small and medium projects. Just ask Sun, HP, various Linux distros, and more listed here: http://wiki.docbook.org/topic/WhoUsesDocBook

You might also find Norm's blog on DocBook vs. DITA interesting: http://norman.walsh.name/2005/10/21/dita

We are also working on some exciting developments for the Interoperability Framework, so stay tuned!

Monday, April 14, 2008

DocBook vs. DITA: revisited

The Content Wrangler has published a very interesting article by Dick Hamilton on choosing an XML schema.

I get asked very similar questions all the time! I think I'll start sending folks to this article as recommended reading...

Wednesday, April 09, 2008

Farewell, Young Arius

One of the greatest, Oscar-winning actors of our times has passed away. Godspeed, Mr. Heston!

Charlton Heston starred in many of my favorite movies, chief among them: "Ben Hur", "The Ten Commandments" and "Planet of the Apes". He was in over 125 movie and TV productions, so he was very active.

My wife had the honor of meeting him at a CBA convention. You'll have to ask her about it!

I also appreciate the amount of work he did for the NRA and other conservative organizations. He will be missed!