[Octave-bug-tracker] [bug #49022] "demo light 11" results in gnuplot errors

[Dan Sebald]

Follow-up Comment #2, bug #49022 (project octave):

Binary gibberish is the data coming through the pipe.

I think this may be a duplicate of the bug Rik pointed out in a similar bug
report:

https://savannah.gnu.org/bugs/?42561

The line for which we put in uint8() as a temporary fix is the source of this
as well.  I make the following change:


diff --git a/scripts/plot/util/private/__gnuplot_draw_axes__.m
b/scripts/plot/ut
--- a/scripts/plot/util/private/__gnuplot_draw_axes__.m
+++ b/scripts/plot/util/private/__gnuplot_draw_axes__.m
@@ -1010,8 +1010,9 @@ function __gnuplot_draw_axes__ (h, plot_
                 ccol = 255*ccol*[0x1; 0x100; 0x10000];
               endif
             else
+color
               colorspec = sprintf ("lc rgb \"#%02x%02x%02x\"",
-                                   uint8 (255*color));
+                                   uint8 (255*color))
             endif
 
             sidx = 1;


and amongst all the screen garbage see lines like:


color =  0.18286
colorspec = lc rgb "#2f


That sprintf() is failing because the color has only one element, not the
customary three color components.  There is negative numbers in the color
data.  I think this is just simply a case of having to run the color data
through the cmap, as we've recently added elsewhere for patch colors (recall
the green sphere inside a yellow sphere with holes?).  Before fixing though,
I'm waiting on some discussion about the way data is organized for
markers/patches.

    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?49022>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/