show that after transaction is not useable with atomics

[barnabasJ]

[zachdaniel]

@barnabasJ would you mind trying out removing this restriction to see if it works with AshPostgres? By my understanding it should be possible to do after_transaction calls with atomics, so it may be as easy as removing this restriction and adding some tests for it.

[barnabasJ]

@barnabasJ would you mind trying out removing this restriction to see if it works with AshPostgres? By my understanding it should be possible to do after_transaction calls with atomics, so it may be as easy as removing this restriction and adding some tests for it.

I'm working on something, I'm trying to understand the whole atomic_after_action thing the regular Ash.Changeset.after_action and the phases around that to see if we need Todo something like that as well

[barnabasJ]

@barnabasJ would you mind trying out removing this restriction to see if it works with AshPostgres? By my understanding it should be possible to do after_transaction calls with atomics, so it may be as easy as removing this restriction and adding some tests for it.

I'm working on something, I'm trying to understand the whole atomic_after_action thing the regular Ash.Changeset.after_action and the phases around that to see if we need Todo something like that as well