Events after event based gateways should have their durations calculated in a way to match the way they are handled in the engine. Currently, it is possible that a catch event that gets it's start time from a previous node update will leave the preceding event based gateway with a duration of 0. We need to test various scenarios in the engine to make sure our logic matches.

ATs:

• The calculation for event based gateways and their subsequent catch events should match the engine