[Python-apps-team] Bug#550635: mayavi2: [regression] interpolated scalar coloring results in black surface

[Gael Varoquaux]

On Sat, Mar 27, 2010 at 04:41:00PM +0100, Francesco Poli wrote:
> I am sorry for my late reply, but you didn't put me in Cc:, hence I
> failed to see that you commented the bug, until I took a look at the
> bug log on the Debian BTS web front-end...

Cced this time :).

> > I suspect it might be
> > graphics-card related. What your graphics card?

> Quoting from Xorg.0.log:

> (II) intel(0): Integrated Graphics Chipset: Intel(R) 965G
> (--) intel(0): Chipset: "965G"

> with xserver-xorg-video-intel/2:2.9.1-2

> However, I've just reproduced the bug on another box having

> (--) SAVAGE(0): Chip: id 8d02, "Twister KN133"
> (--) SAVAGE(0): Engine: "Twister"

> with xserver-xorg-video-savage/1:2.3.1-1

> Hence, it seems that the bug is *not* hardware-specific.

Most probably indeed.

> > Also, it's most probably
> > a VTK bug, not a Mayavi bug.

> It's possible: I am still using libvtk5.2/5.2.1-11 (currently in Debian
> testing).
> Maybe libvtk5.4/5.4.2-5 (currently in unstable) solves the
> issue...

> Should I test mayavi2/3.3.1-1 with libvtk5.4/5.4.2-5 ?

That would be helpful. It would explain why I could not reproduce. If
that doesn't sort it out, we'll have to seek another explaination :).

Cheers,

Gaël