All good points. Clearly the use case I'm reacting to at the moment is the demo/working-with-in-process-model issue, but having a slick way to chart final-data-collection is very handy too. A checkbox could be the quickest, easiest option to facilitate both usages. I would envision a dialogue box that lets you manually specify the image dimensions/proportions, so users who are exporting PDFs or manually tiling charts across the screen and want them to be the same size, etc, have that option. That'd be a bit more work though -- a checkbox would have the same effect. IMO the default behavior ought to be scale-on-drag & fill whole window. You'd also need a mechanism to auto-adjust text size. Siggy On Tue, Apr 23, 2013 at 12:37 PM, Chris Hollander <[log in to unmask] > wrote: > I'll see your chart and raise you a crazy ggplot2 graph! > > Having the option for publication quality charts is a good idea, since > realistically not everyone knows how to use other software. What about a > checkbox that would enable resizing? Then by default the system would work > as it does now, but if you click the box you'd get a resizable border on > the plot and could drag it as you will. > > > On Tue, Apr 23, 2013 at 12:19 PM, Sean Luke <[log in to unmask]> wrote: > >> On Apr 23, 2013, at 12:11 PM, Chris Hollander wrote: >> >> > Are using chat windows directly from mason really the best idea for >> publications? I would think they're much better off for giving >> demonstrations and exploring the model. If you want to graph information, >> wouldn't you be better off just writing the data you want to a file, and >> then using a "real" plotting system to make the graphs? (Like gnuplot, R >> with ggplot2, or matplotlib/matlab, etc) >> >> Actually, the chart mechanism was specifically developed to make good >> looking charts! That's why it outputs PDF, provides dashing and width in >> addition to color, and has a white background. Behold: >> >> >> >> >> So again, I'm cool with changing the default behavior perhaps, but I need >> to know what you want the default behavior to be *exactly*. >> >> Sean >> > > -- Ph.D student in Computer Science George Mason University http://mason.gmu.edu/~escott8/