template drop behaviour around sharp corners

What i'm trying to model is a small channel for a river.

Point conrols unfortunately don't help in this case, the closes i got is by inserting key stations.

I guess it's a case of working more or less as intended, but not as desired

I can see how that would be difficult programmatically because, if you assume a template is essentially perpendicular to the alignment, the "sharp corner" is extended "out" past the actual alignment. You are expecting that outer/left side to continue going beyond the alignment vertex. I agree there are situations where I might want to see that behavior, but I can understand that programming that logic, and finding a way to distinguish that from the standard behavior could be troublesome.

it should at least work with point controls no?

after all, it works in 2d with multilines.

bit more complicated with added z admittedly.

there could/should be a flag that would let you chose if you want to maintain width as defined in the template (rounded outside corner), or geometry as defined in the axis (sharp corners parallel to the axis)

also on the inside of the corner where things get squashed it's always going to be a problem.

You could have a point control on that outer edge, but what else is that last bit of line controlled by otherwise.

As example...Simple template with 2-12' lanes, each sloping away from the centerline at 2%. You can even control the width with point controls along a feature. When you get to that corner, your centerline goes somewhere else. Even if you had the outer edge controlled by a point to a feature, there is no longer a centerline in that region to take the slopes from.

Actually, how would you define those elevations? Flat from the end of the alignment? Continuing the profile grade along the edge? How would that transition to where the outer edge "picked up" the centerline again?

Multilines are a very different, very old, 2D technology and there is very little design logic to them. A template is a complex design of relationships in 3D, so making that kind of (apparently) simple extrapolation really involved.

You could have a point control on that outer edge, but what else is that last bit of line controlled by otherwise.

As example...Simple template with 2-12' lanes, each sloping away from the centerline at 2%. You can even control the width with point controls along a feature. When you get to that corner, your centerline goes somewhere else. Even if you had the outer edge controlled by a point to a feature, there is no longer a centerline in that region to take the slopes from.

Actually, how would you define those elevations? Flat from the end of the alignment? Continuing the profile grade along the edge? How would that transition to where the outer edge "picked up" the centerline again?

Multilines are a very different, very old, 2D technology and there is very little design logic to them. A template is a complex design of relationships in 3D, so making that kind of (apparently) simple extrapolation really involved.