Re: [Gnuplot-py-users] Preparing for 1.8

[Michael H. <mh...@al...>]

bm...@ca... wrote:
> I'm preparing for a 1.8 package.

It's really great that you are working on this.  Don't forget to add
yourself to CREDITS.txt :-)

> I added the svg patch.
> Furthermore, I need to do the patch in Feisty/Gutsy Ubuntu:
> 
> Index: _Gnuplot.py
> ===================================================================
> --- _Gnuplot.py (revision 299)
> +++ _Gnuplot.py (working copy)
> @@ -574,8 +574,8 @@
>                  % (string.join(keyw.keys(), ', '),)
>                  )
> 
> +        self.set_string('output', filename)
>          self(string.join(setterm))
> -        self.set_string('output', filename)
>          # replot the current figure (to the printer):
>          self.refresh()
> 
> to prevent pictures being overwritten with the next hardcopy command. Anybody
> sees that too (eg in new test.py I write a .svg file after ps, and this
> overwrites the gp_test.ps file too with the svg file without above patch)?

The change seems reasonable.

> As Michael said, we should try to keep the test suit working. I'll look into the
> errors I get when time. These errors are present in 1.7 on my box too however.
> Should somebody have time to look at them or have pointers, let me know. Or
> should they not stop a 1.8?

I strongly suspect that the problems shown below are caused by changes
to gnuplot itself.  Several gnuplot commands and options have changed or
been marked deprecated, and Gnuplot.py hasn't been brought up to date.

Of course it would be nice to bring Gnuplot.py up to date with the
current status of gnuplot, but certainly not at the risk of your energy
running out and no release being made at all!

> Doing python test.py gives errors:
> 
> *  test hardcopy:
> 
> Testing hardcopy options: enhanced=1
> Press return to show results...
> 
> gnuplot> set terminal postscript default enhanced
>                                          ^
>          line 714: extraneous argument in set terminal postscript
> 
> *   test GridData and compute_GridData:
> 
> gnuplot> set parametric
> gnuplot> set data style lines
> gnuplot> set hidden
> gnuplot> set contour base
> gnuplot> set xlabel "x"
> gnuplot> set ylabel "y"
> gnuplot> splot "/tmp/tmpdFzv2L.gnuplot/fifo" notitle
> Same thing, saved to a file
> Press return to show results...
> Notice: Cannot contour non grid data. Please use "set dgrid3d".
> 
> gnuplot> splot "/tmp/tmpvEBqT_"
> Same thing, inline data
> Press return to show results...
> Notice: Cannot contour non grid data. Please use "set dgrid3d".
> 
> gnuplot> splot "-" notitle
> The same thing using binary mode
> Press return to show results...
> Notice: Cannot contour non grid data. Please use "set dgrid3d".
> 
> gnuplot> splot "/tmp/tmpcCCPct.gnuplot/fifo" binary notitle
> Same thing, using binary mode and an intermediate file
> Press return to show results...
>          line 1903: File doesn't factorize into full matrix
> 
> Well, that's all for now folks.