On Wed, Apr 18, 2012 at 2:18 PM, Werner LEMBERG <w...@gnu.org> wrote: >> FT_ORIENTATION_NONE is supposed to mean CANNOT BE DETERMINED, which >> is easier said than done. Any contour has a well defined orientation >> as a sign of the area it covers: plus or minus. It is only the >> degenerate contours or the contours with self-intersections that can >> have zero coverage area and, therefore, undetermined orientation. > > There is a difference between `cannot be determined' (by the used > algorithm) and `is undetermined' (in the mathematical sense). I think > FreeType only refers to the former, even if the implementation doesn't > work as expected.
Can I rephrase this like this? It is not the purpose of FT_Outline_Get_Orientation to diagnose broken outlines by detecting sub-contours that intersect or have inconsistent orientation. Therefore, some very horribly broken outlines will indeed result in FT_ORIENTATION_NONE, others will remain undetected. I sense that Behdad has embarked on the mission to implement a thorough diagnostic tool, which is a noble and very difficult journey, which should be separate from fast and dirty FT_Outline_Get_Orientation. Alexei _______________________________________________ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
