XAPKOHHEH

I've been using userData to store some asset state info. UserData was chosen instead of a hidden parameter because userData can be set on a locked node, and should be saved to hip file or HDA just as a parameter.

And it all works... except it turned out that userData on locked nodes is not saved to the hip file. So an asset that consists of assets that store part of the state in userData will not store it's state to hip file, and on each scene load the userData on internal nodes of the asset instance will be reset to what it was during asset creation.

My only question is - is this the intended behavior for userData?

many thanks)

Sorry if my question is not too smart - why not use custom force field in simulation? - it's easy to archive whirlpool effect along any curve this way.
Simulate lowres and then use that velocity fields for highres details?
Or is simulation out of question at all?