Re: Overlap removal
Kristof Bastiaensen <[email protected]>
| Newsgroups | gmane.comp.fonts.fontforge.devel |
|---|---|
| Message-ID | <[email protected]> |
On 27-06-14 00:44, [email protected] wrote: > On Thu, 26 Jun 2014, Kristof Bastiaensen wrote: >> Feel free to add comments, clarify stuff, make suggestions, etc... >> Of course right now it's still incomplete. > When you build your graph, you'll need to find the ordering of the edges > around each vertex; these aren't pure graphs, but combinatorially > embedded. Note that two Bezier segments may share an endpoint, and have > the same direction at that endpoint, without being the same as each other. > In that case the ordering has to depend on their curvature; you can't just > break ties arbitrarily because it may lead to inconsistencies at the other > endpoint. > Thanks Matthew, that's a part I didn't write up yet. I did find a solution, but it didn't involve the curvature. Using curvature may be the easier solution, since it's easy to calculate at the endpoints. When checking if two directions are the same, there may be some error leading to inconsistent results. A good question is how large the tolerance should be, so this can be done robustly. ------------------------------------------------------------------------------ Open source business process management suite built on Java and Eclipse Turn processes into business applications with Bonita BPM Community Edition Quickly connect people, data, and systems into organized workflows Winner of BOSSIE, CODIE, OW2 and Gartner awards http://p.sf.net/sfu/Bonitasoft