The tmax_pseudo_ellipse() fn now handles the zero points correctly.
Authored by: bugman 2010-08-10
Parent: [r11461]
Child: [r11463]