I’ve just started going through the papers from XML 2006. I wish more people were putting their papers or slides up, so many are missing. Now that we have this new thing called the WWW, the readership for a paper is much more than just the conference participants; people like me on the other side of the world who weren’t there would love to read many papers. (I’d love to read the papers from Kitsis, Champion, Kay and Melton for example, who I all admire tremendously.) So here’s my fake realtime blog of the first day’s proceedings…

Michael Hahn. Peaceful Coexistence: The SGML/XML Transition at Cessna Aircraft
“Unfortunately, the SGML toolsets we’ve used to develop our processing system are being supplanted by XML tools and techniques.” “The primary purpose of validation in our current system is to guarantee consistency and predictability for the latter stages of the processing chain.” “We are …more interested in content validation, not structural validation, once the data leaves the author’s hands.” “Content validation is currently done using Perl and pattern-matching, and at some point in the future we may consider shifting these processes to Schematron as our data becomes 100% XML-complient.” “We are right out there on the trailing edge of technology!”

You can actually run Schematron on SGML as well as XML. (Topologi has an add-on for this, which is an ESIS-to-SAX converter.)

Tony Coates. Agile XML Development: Build Environments for XML
Tony looks at JUnit and UnitNG and finds they come up lacking. He has a free Ant task for wildcarding files. This is a good paper for a requirements list for what a Schematron implementation should provide, especially for batch work.

Peter Meirs. Creating Intelligent Workflows with PRISM Metadata
PRISM is a metadata standard that adds all sorts of publishing metadata, enhancing Dublin Core.. Four myths of metadata: “I can find everything I need with a full-text search” “Industry standards have too many elements that I just don’t need.” “One XML vocabulary is as good as another.” “There is no way to show ROI on metadata investment.” Hmmm, without that third myth, the second myth is absolutely correct IMHO: in fact, adopting an industry standard schema without critically evaluating it and having a subsetting process in place is the fast road to hell and project failure.

Webb, Harvey, et al Integrated Public Alert and Warning System (IPAWS)
There is an OASIS group on Emergency Management. The PDFs seem to be about current plans rather than implementation experience. On the other hand, if anyone should be presenting “Masters Series” it is definitely people like Betty Harvey, Chet Ensign and Eric Severson, who have each have enormous technical credibility from decades of experience with markup and sensitive-information systems. I’d have liked to have seen this one.

David Carver. Agile XML Schema Development
The Agile thing is very interesting. It has a lot of similarilities with the XML Extensibility Manifesto. Carver’s idea seems to be to position schema maintenance in an automated enviroment with metrics and regression testing to see what breaks. Similar idea to what Topologi has been doing for some customers. The slides are too terse to make more comments: to all the people who say “Make slides minimal” I say “phooey: if there are no accompanying notes, then the slides should be more detailed than just what is nicest for auditorium readers.”

Claudia Jiménez-Guarin et al Software Construction for Evolving Systems with incomplete Data Definition
The idea unrderlying this paper seems to be that for some kinds of systems, specifically evolving ones, the schema is not an thing that should belong to adminstrators or developers but is something that domain-experts need to be able to add too. As they add data, they need to be able to add fields too, for example. This is pretty much the same point as Schematron abstract patterns: for some kinds of systems the point is not “How do I make a grammar (or XPath!) to capture my constraints?” but “How do I directly represent my constraints with necessary agility?” The paper is useful for showing that there are significant cases where we need semi-structured schema languages as well as semi-structured data.