Bug 711370

Aside from making sure that we are all doing the same thing (and thus presenting a consistent experience for users), and can quickly and easily communicate new ideas and ways of doing things using a common conceptual vocabulary, another important reason for investing a few cycles in getting the syntax right is that it encodes our workflow as a functional specification for the graphical Content Spec assembly tool.

Data export to the tool initially is via this: https://bugzilla.redhat.com/show_bug.cgi?id=710974

Pretty soon we'll need to export an xml representation of existing topic relationships, rather than simply flat cvs.

A flat csv representation is the equivalent to what we're using now, since it's greenfield development. Once we get beyond our first release, however, the tooling will need to be able to represent the existing relationships.

It also needs two modes - one for html (where all topics exist in isolation and appear in multiple places as links) and one for pdf (where each topic exists in one place, and appears in all others as a link). We're targeting the html use case first.

The current contender for the graphical tool portion is: 
Teiid Designer:http://www.jboss.org/teiiddesigner/screenshots.html <- uses GEF
Eclipse Graphical Editor Framework http://www.eclipse.org/gef/

Isaac is working on a graphical interface for publican with the JBDS guys, and they reckon the Teiid Designer is the best starting point, so I'm looking at that.