On Fri, 2014-12-12 at 15:56 -0800, Dick Steffens wrote: > On 12/12/2014 03:43 PM, Tim Wescott wrote: > > This is a medium-bizarre question, but an answer would be of great help > > to me. I've actually asked it on the Scilab list, but if there's not a > > Scilab answer to it, I'd be happy with a Linux one: > > > > I have some papers that I maintain on my web site, for example: > > http://wescottdesign.com/articles/Sampling/sampling.pdf. > > > > These are authored in lyx, with some figures generated with Scilab. The > > site is archived as software, and built using a makefile, including the > > pdf files. Rather than keeping the figures as generated graphics files, > > I keep the Scilab files and generate the figures as needed. > > > > To generate a figure, make runs its generating script from a shell, e.g. > > > > scilab -nw -nb -e > > "execstr(['errcatch(-1,''kill'')';'scf';'exec(''motor-PD-friction.sce'');';'quit'])" > > > > Scilab thinks that it's an interactive environment, so when the script > > makes a figure, Scilab opens the window on top of whatever is running, > > draws it, then closes it. Since I have several papers on the site (and > > its growing), this means that I can't leave the make running in the > > background and get work done, because I'm constantly getting windows > > created in my face. > > > > It's kind of like trying to read in the same room as a cat, except that > > Scilab figures are not warm and fuzzy, and they do not purr. > > > > Scilab does have a "don't use graphics" mode, but if you try to make a > > graph in that mode it bombs. > > > > Is there some way of running a command from a shell that gives the > > command a working X environment (so that it can make the figure), but > > hides that environment from me (so that I can keep designing a circuit, > > answering my mail, or whatever it is that engineers do)? > > > > Thanks in advance. > > > > Just a popped into my head idea, but could you run it in a different > workspace? > > That's definitely a workaround, but it's running right now and it's a GOOD workaround.
-- Tim Wescott www.wescottdesign.com Control & Communications systems, circuit & software design. Phone: 503.631.7815 Cell: 503.349.8432 _______________________________________________ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug