 | | Re: *** Start celebrating .... NOW
|
|
(...) If MLCAD uses Bézier curves to model the bent hoses, you may find some information on the internet about this technique: An interactive demo: (URL) is a more mathematical approach: (URL) may also like to see my own description of how I handle (...) (22 years ago, 25-Dec-02, to lugnet.cad.mlcad)
|
| |
| | Re: *** Start celebrating .... NOW
|
|
(...) ldraw mode program to create flex hoses in the past with no problems, it's just this new way that is confusing me :-) Jennifer (22 years ago, 27-Dec-02, to lugnet.cad.mlcad)
|
| |
| | Re: *** Start celebrating .... NOW
|
|
(...) I think the most confusing aspect of the existing curve program, with the exception of Fredrik's, is the fact that the control points are tricky to position correctly. Fredriik's LDraw mode solves this by dynamically determining the control (...) (22 years ago, 27-Dec-02, to lugnet.cad.mlcad)
|
| |
| | Re: *** Start celebrating .... NOW
|
|
(...) If memory serves me well, in Fredrik's Ldraw mode you specified the endpoints in terms of orientation as well as coordinates, stated the number of joints and it calculated the rest. Certainly a good system for many instances as that (...) (22 years ago, 27-Dec-02, to lugnet.cad.mlcad)
|
| |
| | Re: *** Start celebrating .... NOW
|
|
(...) Some progress - myself and a CAD expert mucked about a bit until we realised the obvious thing I was missing, the fact that the control point coordinates are specified relative to the end points and not in absolute coordinates! Jennifer (22 years ago, 28-Dec-02, to lugnet.cad.mlcad)
|
| |
| | Re: *** Start celebrating .... NOW
|
|
(...) This sounds like a shortcut in the algorithm. If the bezier function is f(t), it does not have the property that |f(t) - f(t+d)| = c for all t, d given. I've corrected for this in my algorithm by using some simple optimization to find a (...) (22 years ago, 28-Dec-02, to lugnet.cad.mlcad)
|
