Hi guys, I've been watch the bullet masterclass from the SideFX website and a question came up while watching the copy to point assembling option.
From what I understood and correct me if ‘m wrong, when you check the “Pack geometry before copying” option, all the geometry is referenced to a single place in memory.
My question is what happens, in terms of “Houdini’s way of thinking” when you fetch two geometry nodes into a switch before and stamp it's value. Does Houdini pack all geometry as one or is it intelligently gets the stamp and creates two indexes in memory for each component feed into the switch.
Can you clarify this to me?
Happy learning.
Thanks.

Satamanster
Hi guys, I've been watch the bullet masterclass from the SideFX website and a question came up while watching the copy to point assembling option.
From what I understood and correct me if ‘m wrong, when you check the “Pack geometry before copying” option, all the geometry is referenced to a single place in memory.
My question is what happens, in terms of “Houdini’s way of thinking” when you fetch two geometry nodes into a switch before and stamp it's value. Does Houdini pack all geometry as one or is it intelligently gets the stamp and creates two indexes in memory for each component feed into the switch.
Can you clarify this to me?
Happy learning.
Thanks.

There's an option on the copy SOP to toggle that behaviour (cache stamping geometry).