The following entry
has brought to light an interesting issue about the clustering process in GH2. The issue of interest is the manual process involved in the clustering in a GH definition. From my experience, this process can be a bit tedious. The following example illustrates this perception.
spring before clustering.gh (17.4 KB) generates two sets of springs with different diameters. As shown on the GH viewport, they are 2 sets of components, one for each set of springs. In order to have less components and less wires, it is usual to create a cluster “spring” that will factor both sets of components into one component “spring”.
Let’s factor the first set
In my opinion, two issues can be considered
- the cluster have duplicate entries.
- replacing the second set of component by a new cluster instance will be a bit hairy (lot of plumbing work ahead)
This is where a “classification yard” component could be usefull. To create a “classification yard”, you first select a set of wire. For instance by dragging the mouse from right to left with the Right Mouse Button pressed as show in the following figure.
Then,
- for each wire selected from one component to one or more component, there is one input and one ouput in the “classification yard” component
- the order of the input and output of the “classification yard” can be edited by the user for visual purposes (“detangle”)
Below is a mockup. The group below would be the “classification yard”.
Notice that this component is purely visual (not creation/transformation of Rhino objects)
If this component is available, I surmised this would mitigate both issues stated above and therefore ease the “plumbing” part of the clustering process.
Clustering could be done the following way
- For each set of component that will be clustered or replaced by a cluster, select all incoming wires and create a “classification yard” component. Do the same for all outgoing wires.
- cluster the first set of component by including the input/output classification yards
- for each set of component to be replaced by the cluster, copy the cluster near the input/ouput classification yards of the set
- do the “plumbing”