Is there any way to convert workitem execution from dynamic to static state?
I'm using a static generation than dynamic and I would like to convert execution to static back again.

Static work items for the whole graph must be known ahead of time, before any of the work items cook. You can use a partitioner node to create static partitions that end up depending on dynamic work items later on. I've attached a simple example of that.

Starting with the next release, it'll also be possible to create a processor node that is dynamic, but also waits for all input dynamic items to be generated before it creates work items. It behaves a lot like a static node does now, but also works with dynamic work items. It's not possible for us to backport that to 18.0 because it required significant architectural changes.

It will be useful!
Thanks!

because right it is still not possible to wait for an attribute or result data from the dynamic workitem (as I understand from your example) if I need static in the end