spiro splines will convert to castel-curves, but interactive editing will retain only control points for those spiro handles.
All features of spiro spline should be implemented. G2 curves are preferred over G4 though. That includes lineTo points which will be useful when interpreting the segue from casteljau to spiro on the same anchor.
More research needs to be done to understand how curvature is passed into the integrator. According to Raph, the curvature of any curve could be passed into the "black box" and g2 continuity will come out. That means that, in theory, a casteljau anchor complete with pole could lead into a spiro anchor which could lead to another casteljau anchor and pole on the other end.
refactored features
[ ] one-way anchors (handle these as chain-breaks? handle these as re-assigned get_th functions?)
[ ] closed shapes (no idea how this is done yet)
extra features
[x] find the location of the cosebase that controls angle
[x] clean up seg_to_bez to return just one segment
[ ] use SVG arc when circumscribing only 3 points instead of passing through the integrator
[ ] clean up entire code base so that the spiro algorithm accepts one Curvemixer-style Chain object and returns completed SVG data
wraybowling
commented
10 years ago
spiro splines will convert to castel-curves, but interactive editing will retain only control points for those spiro handles.
All features of spiro spline should be implemented. G2 curves are preferred over G4 though. That includes lineTo points which will be useful when interpreting the segue from casteljau to spiro on the same anchor.
More research needs to be done to understand how curvature is passed into the integrator. According to Raph, the curvature of any curve could be passed into the "black box" and g2 continuity will come out. That means that, in theory, a casteljau anchor complete with pole could lead into a spiro anchor which could lead to another casteljau anchor and pole on the other end.
refactored features
extra features