For discussion sake the software also fails if one is using two SIMPLE assemblies from Generic links; it isn't that it fails when the assemblies are 'complex' it fails consistently, to both project the sampled lines at the command level frequency settings, and it becomes disconnected from the baseline trough arc segments.
This will be my last post on the subject- but you claim that it's a defect because help tells you that an assembly is always tied to the baseline- but what help means is that the subassemblies bits always remain attached to the _assembly baseline_ that vertical piece on an assembly that hooks it all together
not the alignment you specify as the baseline. Also, the command setting sets the default region frequency- it never says anything about it being a baseline frequency because there is no such thing.
I'm not just making this up as far as how the software logic works- I spent a lot of time with the product designers, developers and QA when writing the assembly and corridor chapters of those three books that I wrote that you hate so much. They aren't perfect books (don't want to talk about that again, many threads already on that subject) but the explainations of the program logic for corridors is very sound and straight from my research. It isn't just something I made up.
I'm sorry it doesn't work the way you want. I'm messing with my stream some more and it seems in order to do the transitioning properly, I have to whip out my calculus to figure out the curves. I will probably do it since I am a geek that way and b/c i am curious if something like that could be done by the program or if there are far too many variables.