18 Jan
2005
18 Jan
'05
8:36 a.m.
Wookey wrote:
Hmm. I just checked. I found one in goon_s1 that the wrong way round, but every pillar in goon_s2 (which is the one coming out wrong in the final plot) is correct (tick on the inside).
The line orientation in inner pillars is significant only for displaying of asymetrical walls (blocks, sand, ice), not for determining the inside of the passage. You have actually two crossing scrap borders in one scrap :-) MetaPost didn't notice it, which is really bad. This can't be improved at the level of MetaPost macros, because we use low-level built-in function to determine the line orientation. The enclosed picture shows the problematic connections. M.