i agree with Mchael on this issue. The suggestion of hiding the input objects is exactly what i do not like in other programs such as Amapi and Hexagon. Then how can you reference one input object to many different ouput objects? If you want to see what is more interesting for history, take a look at SolidThinking. They have a full history tree, while the objects remain individually but also linked through the tree of all objects which used it as input...

jonah

Hi Marc, yeah I think there is some potential to make this work through a kind of layer or grouping type system.

If you could assign the input curves to be grouped with the output object, and therefore move together with them when you repositioned the object, but have them hidden, then it seems like that would have all the ingredients in place.

I'll try to consider this scenario when I work on grouping for v2. Sometimes grouping with hidden objects can be tricky itself though.

- Michael