NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Hello Oscar, I am very glad to see you interested in this. > Is there a design plan for adding graphs to visad? Well, I am contemplating to do so, but I do not have a real plan yet. I am eager to discuss this. > I am curious how you > plan to factor out the layout from the rendering of graphs, given that it > is so tied to domain-specific constraints. I plan to do it in a way, that the rendering is not concerned with the layout, it receives data from the layouter(s) and maps it to DisplayRealTypes. (I would like to stick to Visad's existing renderers.) I am pretty new to visad, so I wonder if you are concerned with performance of such a probably naive approach. > Will there be a set of common > graph-layout algorithms as well as a java interface for providing a layout > algorithm? Hopefully. In the next days I will try to make the simple but effective TouchGraph (http://touchgraph.sf.net) layout code work with a visad 3d example, just to get a feel for it. My final goal is to have a linear arithmetic constraint solver (like the great Cassowary http://www.cs.washington.edu/research/constraints/cassowary/) play along with various layouting algorithms concurrently. This would allow for an interesting mix of strict and loose layouting strategies - and probably it's a benefit for the design because this requires a very abstracted notion of the domain objects, so the systems (graph layout, solver, visad) can play together. I will post on this list, when there's something working. If you have ideas please, by all means, post them, so we can discuss. Yours, Manuel
visad archives: