Re: Octave/OS X Aquaterm problems (bugs or me?)

[Per Persson]

On Jan 7, 2004, at 23:00, John W. Eaton wrote:

> On  7-Jan-2004, Per Persson <address@hidden> wrote:
>
> | the modified figure.m for OS X _should_ work if you set the default
> | gnuplot term to aqua.
>
> If the figure.m file for OS X is different from the version
> distributed with Octave, can you please submit a patch that will
> integrate whatever changes are necessary for OS X with the Octave
> sources?

diff gives (save for the missing docs):

--- /opt/local/share/octave/2.1.50/m/plot/figure.m      Tue Jan  6 
17:11:28 2004
+++ /opt/local/share/octave/2.1.50/site/m/octave-forge/MacOSX/figure.m  
Tue Dec 23 22:26:45 2003
@@ -39,7 +39,11 @@

   if (nargin < 2)
     if (gnuplot_has_frames)
-      if (! isempty (getenv ("DISPLAY")))
+      if (! isempty (gnuterm=getenv ("GNUTERM")))
+       oneplot ();
+        figure_list = union (figure_list, f);
+       eval (sprintf ("gset term %s %d\n",gnuterm, f));
+      elseif (! isempty (getenv ("DISPLAY")))
         oneplot ();
         figure_list = union (figure_list, f);
         eval (sprintf ("gset term x11 %d\n", f));

I'm not sure however that this is the best, or even a good, solution to 
the problem that OS X users might want to use aquaterm as well as X11 
with gnuplot and still be able to use figure(n) from octave.
If only X11 is used on OS X, there is nothing to change in figure.m

To me it seems like a good compromise that the file is distributed with 
octave-forge and only installed on OS X.

> If there are other files that need modification to work better for
> OS X (or other strange operating systems like, say, Windows) then I
> think it would be best to also do the same for those files rather than
> having modified versions of files floating around.

This is also part of octave-forge but may be OK for inclusion in octave 
base.
The changes are not necessary for it to work on OS X, but makes the 
behavior more mac-like.
I've attached a patch.

/Per

image.m.patch
Description: Binary data