> > My thoughts were that a CGI interface could control what the Workspace might > > control. So everything in the Workspace that would require user input, would > > instead bring up an HTML form. And everything that would return an figure, > > would convert the figure to GIF or JPEG and make an HTML page for it. > So it is important that viewers (and possibly the figure builder) be able > to not only present to the screen, but present to paper (postscript?), and > present to graphics (PNG or JPEG). Of course ghostscript will convert > postscript to PNG,JPEG,... If all viewers render to a gnome canvas, can't we just dump that to a file? I think the gnome printing api will even allow rendering to memory without opening an X window (I sure hope so).