Because we need to do batching, regardless of if it is a less explicit Future approach or something like described in #23, the requests need to happen asynchronously from when they are queued up. This means that the stack at the time the request was queued up is lost. This stack is useful however if something is wrong with that request.
We should capture the current stack when requests are queued up, and if they fail, we can present that stack along with the stack trace of the failure.
Because we need to do batching, regardless of if it is a less explicit
Futureapproach or something like described in #23, the requests need to happen asynchronously from when they are queued up. This means that the stack at the time the request was queued up is lost. This stack is useful however if something is wrong with that request.We should capture the current stack when requests are queued up, and if they fail, we can present that stack along with the stack trace of the failure.