|
From: Ethan M. <eam...@gm...> - 2015-07-10 19:41:25
|
On Fri, Jul 10, 2015 at 11:53 AM, Thomas D. Dean <to...@sp...> wrote: > On 07/10/15 11:34, Ethan Merritt wrote: > >> >> A pointer to the discussion and proposed fix would be sufficient. > > > > discussion, among others: > > http://comments.gmane.org/gmane.comp.graphics.gnuplot.user/7933 That was a different problem altogether. The executable was installed in the wrong directory. > http://gnuplot.10905.n7.nabble.com/Reworked-qt-terminal-is-much-faster-and-fails-to-work-on-Mac-td17949i60.html > > Daniel J Sebald proposed a patch. > > https://sourceforge.net/p/gnuplot/patches/653/ > > for cleaning up the linking and waiting for gnuplot_qt socket > > I am not sure this patch is a fix. That patch was a non-starter because it totally broke on OSX and basically just papered over the symptom by not checking for a problem with the font request - so no error message but also the font + scaling was still messed up if the system didn't respond in time. I remain dubious that this is really a problem on the gnuplot end. The font request really is delayed or late, which points to an issue with font configuration at the operating system level. The identical gnuplot code works fine on other linux machines. That doesn't rule out the possibility that something could be done in gnuplot to reduce the chance of triggering an error in the font support, but so far I have not seen a plausible suggestion as to what that change might be. Out of curiousity, are you using Qt4 or Qt5? Which exact version? |