Hi Pilou,
> if I take the same orientation of the orient function
> than my object that works on the top curve surface,
> that don't work for the single curve!
You mean snapping on to the edge curve instead of snapping on to the independent curve object?
Yes, that's normal to get different behavior between those cases - the edge curve knows that it is part of the surface and anything that snaps on to a surface aligns the target z axis frame to the surface normal. When you snap on to a curve that is not part of the surface it will instead align the z axis to the curve tangent.
> That is curious that we must take some different regulate for a
> curve who have same position except a move displacement
The difference is that one belongs to a surface, and so the direction is aligned to the surface - the other one that does not belong to the surface does not have a surface normal to align to and instead aligns to the curve tangent.
> In the file , I want move my object like your cone :)
Just put the view into a similar direction as what I shown above and then it will behave like that - here is an example with your object:
Is that what you want?
But what is the final position that you want this object to have? If I knew the final result that you were looking for it would be easier for me to give you some steps that would give you that result.
> I must find a tricky way for not have hesitate to choose
> the good orientation of the axe of the Orient tool!
Don't worry so much about what things look like while you just move the mouse around - just focus on the final spot where you wish to place the object, and if the position in that final place is reversed from what you need just click this checkbox here to flip it:
- Michael