Has there been any discussions about the possibility of rebuilding a component, or entire rig, after the character has been skinned? Sometimes joint positions need to be changed. Or new components added to an existing rig.
I have an idea of how I would like to accomplish this. But I thought I would source the forum before I start looking at how much the existing framework would need to be altered. Or if it is possible at all.
Hi @izze
Yes this is possible. You export a skinPack (I recommend ascii rather than binary so you can edit the text files) to save your skinning, and when you rebuild, you import the skinPack. There are menu commands, as well as Python commands to do this.
The framework doesn’t need to be “altered”. The ability to rebuild the rig from data is pretty much the point of Shifter and mGear. 
Thank you for your reply!
I’m thinking more along the lines of rebuilding the mGear rig without the need to also rebuild the skincluster. I have a secondary joint based deformation rig that constrains to the mgear skeleton. I also have a F.A.C.S. rig that utilizes the mGear skeleton. Rebuilding the mGear rig, requires me to also rebuild each secondary system.
My secondary deformation rig can be adjusted, or rebuilt, on the fly without the need to rebuild the skincluster. And since all secondary systems utilize the mGear skeleton, I hope to find a way of keeping the mGear skeleton in the scene when rebuilding components.
The most common scenario I encounter is when the modeling department receives a client request to modify the characters proportions. The ideal solution would be to adjust mGear guide positions to match the new proportions. And then rebuild without effecting any secondary rigs or skinclusters that rely on the mGear skeleton to exist in the scene.
A overly simplified example of a solution would be:
Maybe someone would be interested in a collaboration?
One way to solve this problem is to focus on creating a clean interface between the mGear base rig and the additional secondary systems. For example, for a head component, look for every place that requires a connection to the head transform and connect them all to a single transform node. This single transform node can can then be connected to the mGear rig or disconnected when swapping rigs. Similarly, send the deformation to the output mesh with a single blendshape. By cleaning the structure like this, the head component can be plugged into the mGear rig by making just two connections, and the base rig can be deleted and rebuilt without causing issues for the head component.
Creating similar clean interfaces for your add-on components means that they can each be added with a few lines in a simple post custom step.