Writers just wanna write. Why is it so hard to use a structured editor? Keyboard controls, metadata, insertions, formatting controls, moving blocks, references. They narrowed to tasks like creating new document, add a new list, add a reference, etc. Added ‘new’ and ‘new-doc’, outline, mark, crossref, cite. Measured where people got it right and wrong. The new button works. Splitting and joining didn’t. Moving text blocks needs much better context sensitivity and navigation. References work better when the target list is in the document pane. Smart insertion: never refuse to paste, make it fit. Target markup adoption: make all pastes adopt the surrounding styles of the target location.
Graph representations of structured documents. Trees and XML documents interchangeable. Overlap problem, like verse and sentence in poetry. He seems like Danny Kaye playing a markup expert with a French accent. Overlap only TexMECS. Trivially extended MECS. He gives no time to understand his far out ideas, just gives the mood of the story of the paper, I guess. Winner of the Golden WTF Award.
Documents need indexes. Slowness kills. Lux makes queries faster.
PubFactory, online publication platform. Uses XQuery for search. Recently working with Solr, text search index, NoSQL document store, query language. No markup awareness. Lux is Saxon plus Solr. Uses Lucene syntax for search queries. Optimizes queries.
Web as database. Minimize costs. Be user focused. Build for the future.
The web revolution. Provides global person to person communication. Cheap to use. Works as REST system. They can generate XForms based on XSD.
There is a RelaxNG DTD compatibility specification that allows for the default values that DITA needs. They implemented it in Xerces. This makes RelaxNG look more attractive for DITA than DTDs. XSD is altogether unsuited for DITA. DITA 1.3 will use RelaxNG.
XML world domination. But not preferred for authoring. Make an anything parser with a syntax description,variant of vwg format… Can round trip. Presents the syntax description in fairly complete detail.
The big xml tools have hardly changed in the last ten years. Tools elsewhere have changed enormously. Personal anecdotes about where web development has thrived while the markup community has been relatively subdued.
This raised some (polite) hackles. One respondent remarked that while the web people came and went, the core of the markup community has always been the document people, and we’re still here. So I realize that I’m a document person in just this sense. But my recent interest in the IndieWeb means exploring a different world. I already sort of knew that, but this underscores that.
All XML documents can be considered as one forest. The info space is the sum total of all accessible XML nodes. The space image extends the navigation model. Space potential, boundaries, and limitations. Physical, format, and API boundaries. Asserts that the JSONic approach abandons the Info Space idea. Proposes more structure between trees than URIs.
Amazing visionary philosophical presentation about expanding the potential of the XML world.
Semantics don’t equal values. Semantics are about meaning, values are about presentation. Different profiles can happen to have the same values. Localized profiles and versioning cause different values. Use an abstraction layer. If the core semantics change, the values may or may not change. Use urn’s. (This might be useful for airplane effectivity?) Very detailed listing of the practical possibilities, as I’m starting to see is usual for Ari.
Publishing: packaging and distributing of ideas. XML single source semantic markup (DocBook) and automated transforms. From HTML5 to PDF via antenna house CSS! DocBook tool chain very heavy. DocBook not ideal for digital first content development. Common thread of current output formats is HTML and CSS. Not all authors love writing in DocBook. Things like Medium are the future of publishing. O’Reilly Atlas: wysiwyg, web based, responsive, version controlled, seamless. HTMLBook! A DocBook-ified HTML5. With XSL tool chain.
People are really buzzing about this one. Some predicting this is a bad move, others excited.