so far so good. It's a bookkeeping-problem
The latest, and so far best image, had no flickering triangles, but
the redundant part of the data used for the vertices were stribed.
With the redundant part I mean a rather large square grid of equel
size and (almost) equal angle triangles. An bunch of values from a D2
array of elevations has been reasigned a value of 0, where the author
originally has used a sort of error-value of -9999. ... how can
something like that propagate into the image as a kind of disturbanse?
Could VB silently pass a NaN-value (deviding by zero), somewhere.
I asign a color-value to those z=0 verteices .. but no colored
vertices has ever shown. Not when things run smoothly, .. maybe
directX is struggling to do it, I mean, tries to understand my
intentions, and not my (possibly erroneous) code.
Before the object is rendered, I asign a material of a different color
to the call. I've deliberately skipped any considerations involving
colors as of secondary importance.
I used to feed colorvalues as like color.red.ToArgb. I may use another
way when loading from file, but something that's perfectly digestible
as an integer. Again, I used to load /all/ shapes to be use in a
session from file, without problem. The color that shows is 'business
as usual'.
Folks, becourse I started that 'serious overhaul' and rewrote a bunch
of stuff - I'll have to consider all options. That's a mess, but I've
only registrated the one problem I keep buzzing about.
Carsten
This computer is old, usually needs more than one start to get going.
Files dissapear. It's my net-computer, and I could theoretically be
compromized from malicious code. I don't have a chance to know. To
spare myself for those considerations, I keep my coding-pc off the
net ... and it's not yet old enough to need glasses for a file-
read ..
Carsten