NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.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.
Hi Vikram, > I'm currently finishing up a report of my experiences with VisAD, and > everything seems to be going ok except for the section on VisAD's use of > Java's Remote Method Invocation (RMI). > > I would like for the readers of my document to get a feel for the > architecture of VisAD's remote-data model; unfortunately, it would seem > that I don't understand it too well myself :-) > > What I'd like to know is whether the rendering of the remote data takes > place on the remote machine. Or is the data ported over to the client > to be rendered there? The second scenario seems unlikely, given that > VisAD was designed to work with such large data sets, but I would much > rather hear from the authors than my make my own semi-educated guesses. The default is that data are rendered in the DisplayImpl, which is generally on the client. The presence of the scene graph in the client supports greater interactivity. However, we are planning to add an option to support a remote slaved display from a DisplayImpl on the server. BufferedImages from the DisplayImpl.getImag() would flow to the slaved display on the client, and MoueEvents would flow from the slaved display back to the DisplayIml on the server, or user interactions. This will not be too difficult to implement. Applications may also construct CellImpl objects on the server to derive data objects of reduced size to be passed to the client for rendering in a DisplayImpl. The data reduction may be under interactive user control. Cheers, Bill ---------------------------------------------------------- Bill Hibbard, SSEC, 1225 W. Dayton St., Madison, WI 53706 hibbard@xxxxxxxxxxxxxxxxx 608-263-4427 fax: 608-263-6738 http://www.ssec.wisc.edu/~billh/vis.html
visad
archives: