Known Camera Poses in SfM?

[contagon]

Been perusing through the code, I had a couple of things I wanted to make sure I understand correctly.

The results from importing known poses with the ImportKnownPoses node will naturally be used in image and feature matching, but I'm curious how this information is used in the actual SfM. From looking at the code, it appears to essentially be used as an initialization for SfM. Is this correct? I just wanted to confirm that these values are still optimized for and not set as constant or anything.

Another, more out of curiosity, is what is a "rig constraint"? It appears to take effect when the number of poses gets sufficiently large, I was wondering if you have any resources/documentation that explains a bit more about it.

Thanks for such awesome software!