How to read copy stamp data flow

[catchyid]

Hi,

Beginner question.

I understand how to do copy stamping, however I am not sure how data flows inside the network, for example (see attached image):

"copy" node needs both "platonic1" and "sphere1" to work, so they should be evaluated first, i.e. before "copy", HOWEVER, "platonic1" references the stamps defined in "copy" node, which means "copy" must be evaluated first!?  My question: which one is evaluated first, or how this network works? 

Thanks,

[Atom]

Stamping can be a little weird to understand when you are first introduced to it. It seems like you are going "up-stream" to achieve the goal. But what is really going on is the Copy node is evaluating both it's inputs. Left input is the object to instance and right input is where to instance the object from the left input. Both platonic and sphere have to exist before Copy can evaluate them.

 

I am guessing at the code flow here but I assume that Copy generates a list of points from the right input then looks up the left input network tree and evaluates any expressions found. The expressions return values which are only valid for that copy instance. Copy then generates an instance at the point location with those values. It repeats that process for every point on the right side input. If there is no right side input then the point list is generated from the parameters in the Copy panel.

 

You may want to check out the example file at this link which compares Copy vs. Instancing and the speed difference.

Edited August 15, 2016 by Atom

[Guest tar]

You can see dependencies by turning on 'Show Local Dependency Links' in the Network Editor Display Options. A coloured flow arrow will appear.

[catchyid]

Thanks guys