We see some conditions where back-to-back test runs from different tools can cause results to be different. It's challenging (in my opinion) to predict how long we'd need to wait universally between any/all tests to avoid problems. So, my proposal is to introduce a pause flag with a configurable time wait between test runs. This will allow test operators to have more direct control over the behavior of bobber, and allow some level of experimentation to discover the "correct" wait time between test runs for a given configuration.
We see some conditions where back-to-back test runs from different tools can cause results to be different. It's challenging (in my opinion) to predict how long we'd need to wait universally between any/all tests to avoid problems. So, my proposal is to introduce a pause flag with a configurable time wait between test runs. This will allow test operators to have more direct control over the behavior of bobber, and allow some level of experimentation to discover the "correct" wait time between test runs for a given configuration.