3.0.4 call for patches

[Ben Abbott]

On Friday, December 12, 2008, at 01:53PM, "Jordi Gutiérrez Hermoso" <jordigh at gmail.com> wrote:
>2008/12/12 Kai Habel <kai.habel at gmx.de>:
>> Jordi Gutiérrez Hermoso schrieb:
>>> 2008/12/12 Jaroslav Hajek <highegg at gmail.com>:
>>>
>>>> In any case, the error message I get looks like an error from gnuplot,
>>>> so I'll try gnuplot 4.3 and report back.
>>>>
>>>
>>> Uh, you get an error message? I don't. I just get a solid blue plot.
>>>
>>> - Jordi G. H.
>>>
>> Blue or cyan? We had an error like this mid-October [1]. Are you using
>> an octave version older than two month? What is your gnuplot version?
>
>Like this:
>
>     http://platinum.linux.pl/~jordi/piccies/Screenshot.png
>
>Using the latest tip and gnuplot 4.2
>
>Btw, I had to completely recompile Octave starting from autogen.sh,
>because I was experiencing several problems. I thought that doing "hg
>pull -u && make && make install" should be enough for me to get the
>latest running version of Octave. Is that incorrect?
>
>- Jordi G. H.
>

That looks like the gnuplot bug related to 4D data. The problem is that gnuplot is mixing the scaling of the z-axis with color data (4th dimension). If memory serves, you can try adjusting your z-axis to verify. Just make sure the zlim encompasses the extent of the colors.

I recall patches being produced for both gnuplot 4.3 and 4.2, and am sure 4.2.3 does not include it (although I did patch my 4.2.3 sources and it worked fine). The patch may have made it into 4.2.4, or it may have just missed the cutoff.

Perhaps it is prudent for someone to download the developers sources for 4.2 and compile to see if the problem is then resolved ... sorry I don't have the time at the moment ... my work and home life are a bit too busy for such leisure activities ;-)

Ben