Hello James,

Good to hear, that you still make experiments:-)
I had a look on your file. There are two problems in my opinion. First - the placement of the circle dosen't fit allways with the startpoint of the railcurve and the second problem is the orientatiion of the u-and v-line. I think the "sweepnode" is difficult to handle and we have to think about, how we can make it easier to use. The name sweep is also not very handy - any ideas? For the orientation it would be also better, if the iCurve-Node would give a frame for start and endpoint, but I didn't find a solution for that at the moment. And I think a native sweep node would also be very useful. I've attached a changed file, that isn't perfect (orientation of the profile dosen't fit) but works in a first step. I hope that you can find a solution based on it. To analyse the problem, I have cloned some objects and gave them to the output - so cut the wires or delete the clone-nodes.

Have a nice day
Karsten