The primary user interface is a window containing sub-windows.
All working buttons (should) have mouse-over hints.
The three most interesting windows provide views of the data at different levels of granularity. The Perspectives window deals with top-level, system- and project- wide views. The IdeaGraph Editor window is used for the diagrams, which are comprised of Ideas (which are the red rectangles) and Connections (the lines between the rectangles). To keep things simple these diagrams are called IdeaGraphs. The Idea window is used to provide the main details for each idea (node) in the IdeaGraph.
Double-clicking on the IdeaGraph Editor window creates a new Idea. Single-clicking on an Idea makes that Idea the focus of the Idea window. Double-clicking an Idea opens the Document Editor window (see below)
The NEW Way Connections are Drawn
When you move the mouse cursor over an idea it will initially change to an arrow. But if you move the arrow over the centre of the idea then the cursor will change to a hand shape. Clicking and dragging now allows you to draw the relationship lines between the ideas.
If a connection (arc) is clicked on, a different window appears in place of the Idea window :
The connection can anything you like, but connections from existing vocabularies may be selected from the menus. Other vocabularies can be loaded using the Vocabs tab in the Perspectives window. Once loaded they will show as additional menus here.
RDF Geek Note : that may be read as RDF Properties from existing ontologies, custom user input is recorded as a property in an RDF Schema in a custom namespace.
User Interface Geek Note: the creation of nodes and arcs is fairly non-standard, but most standard tools use a fixed toolbar. If you hit the F11 key in IdeaGraph you have the full screen in which to work on the diagram. It only takes a few minutes to get used to, and no trip to a toolbar is needed.
Other info is available from the Idea window :
Elephant's Note : In the prototype I had a lot more variation in the views of the nodes in the graph : shapes, images, colors. These will be back in time for version 1.0.
Double-clicking an Idea in the IdeaGraph Editor window opens the Document Editor window :
This stuff can form part of a blog. Etc. See
Creating an Ideagraph - note in particular the IdeaBox, not covered here
Returning to the Perspectives window, it has some interesting (and not so interesting) tabs :
The Project and Channels window display tree for managing projects and syndicated channel feeds respectively.
Geek Note : RSS feeds are pre-processed with XSLT
I messed up a bit of the code here. These will be friendlier!!
The Vocabs tab is for managing vocabularies, both existing ones and ones you have created yourself.
This still needs a lot of work.
Geek Note : I'll probably use another RDF tree component here too. The bit below was a start at separating a friendly version of each properties (rdfs:label) and the internal representation (rdf:Property)
The Settings tab is for project, user and system configuration :
Geek Note : Bit crude at the moment, but all the config stuff is stored in RDF files.
The Command tab is for command-line operations. This is presently rather geeky, though the plan is to also enable simple commands to be entered here to manipulate the IdeaGraphs.
Geek Note : this stuff's pretty cool. Python's built on Jython, Prolog's home-spun.
There are two tabs, Prolog and Python, each corresponding to an interpreter in that language.
Here a Python command is used to display a serialization of the simple IdeaGraph. Existing source files can be loaded and executed using the buttons (they have mouseover tooltips).
Geek note : see what I mean - but it gets better...
This is very much experimentation-in-progress, but currently the Prolog engine has a separate knowledgebase, which can be populated from the data in the IdeaGraph's RDF model. You can then manipulate that knowledge, e.g. filter-by-retraction, and then push the knowledge base back out to another IdeaGraph diagram. There is a file with the IdeaGraph distro (ideagraph\data\prolog somewhere) which contains some of the RDF(S) axioms, but most are commented out because it soon goes into a loop - tabling isn't yet implemented. There are some other comments here.
Non-geek Note : I did warn you!
The buttons have the following functions :
Geek Note : the Prolog interpreter, Java and Jython all share the same VM, so exactly the same object instances can be worked on with any of the three languages. A simple example of this can be seen here
The parts of IdeaGraph in the screenshots so far have been related to core functionality. There are a few things that you might want to do around the fringes that call for external tools. Bare-bones versions of the obvious tools will also be provided :
Features mentioned previously in the (first attempt) Tutorial are still available :
Creating an Ideagraph - note in particular the IdeaBox, not covered here
(there have been a few name changes, and a little window rearranging)
Please Note A Change in the Way Connections are Drawn
That's most of it, for now.
If your eyes have now glazed over and your head is pounding, here's a baby hedgehog to help you feel better :
