[2202] in Athena Bugs
(VS2) Version 6.2A: ico -obj pyramid
daemon@ATHENA.MIT.EDU (John T Kohl)
Mon May 8 10:39:10 1989
Date: Mon, 8 May 89 10:38:37 EDT
From: John T Kohl <jtkohl@ATHENA.MIT.EDU>
To: bugs@ATHENA.MIT.EDU
[feel free to forward this to the X consortium; this is relative to
X11R3+ (our) sources]
System type, version: (VS2) Version 6.2A
System name: lycus
What's wrong:
The pyramid exhibits some problems in the hidden line/hidden face code
in /usr/bin/X11/ico. This is most clearly visible on a color display.
/usr/bin/X11/ico -obj pyramid -faces -colors blue yellow green red cyan -sleep 1
will spin a pyramid around. If you watch carefully, it is easy to see
that the color updating of the visible faces is not being correctly
performed. It appears that the base of the pyramid always ends up 'on
top' instead of being underneath as it should be at times.
This hypothesis is supported by occasional flashing of the colors of the
top faces before getting overdrawn with the base's color.
The other shapes seem to work ok.
An experimental version defining a tetrahedron as follows causes even
more bizzare 'hidden' line/face breakage, in a similar vein:
/* objtetra.h - structure values for tetrahedron */
{ "tetrahedron", "tetra", /* long and short names */
"tetrahedron", /* long name of dual */
4, 6, 4, /* number of vertices, edges, and faces */
{ /* vertices (x,y,z) */
/* all points must be within radius 1 of the origin */
{ 0, 0, 0 },
{ 0.57735027, 0, 0 },
{ -0.28867513, 0.5, 0 },
{ -0.28867513, -0.5, 0 },
},
{ /* faces (numvertices + indexes into vertices) */
/* faces must be specified clockwise from the outside */
3, 2, 1, 0,
3, 3, 1, 2,
3, 3, 2, 0,
3, 3, 0, 1,
}
}, /* leave a comma to separate from the next include file */
/* end */
What should have happened:
The algorithms shouldn't barf on pyramid-like objects. I suspect there
is some assumed geometry constraint which isn't true for the pyramid and
tetrahedron.
John
