inevitable wrote:the elephant in the room is this: how are you testing your changes?
i think the 'unable to find evidence but please tell me where/when to look' thing is similarly hard to swallow because you haven't specified what you've searched for or what you might be missing.
it's understandable there's some cat and mouse behavior around tracking player actions, but if you can just say, for example, that you searched exhaustively for catas being built this world, as in each cata built has a unique object id or event recording of its construction, and those catas weren't used to damage walls within 24 hours of their creation, then that should remove all doubt, no?
i guess it would be neat/reassuring/something, if you could share how many catas were built this world, and how many were destroyed before they were used, and how old the catas were when they dealt damage to anything.
Loftar and Jorb actually have a regression environment with an entire automated test suite testing various use cases. When loftar is close to finishing his Sprint and he moves the Story into the "Ready for QA" status, he's actually just running this through his build pipeline of millions of unit tests. Once that merges, they create a release branch and push it to Regression where they load test by spawning in 1000 hearthlings at each thingwall in the test environment, and then lastly, Jorb the hero steps in and does some manual QA to verify that the feature is functioning correctly for the end user. After this, Jorb and Loftar hold a retrospective for their team of two people where they talk about what went right, what could have gone better, what needs to change, and kudos! In this meeting, they go over their Say/Do, sprint velocity, any form of churn in the sprint, etc etc.
You're actually really lucky! Because this bug was a Severity 2 incident, Loftar deftly added the logged bug (after triage of course) to the sprint and took out OCO from the sprint (shucks would have been done this sprint too!) so that he could fix this for the community. This was almost as bad as the Severity 1 Incident where we couldn't destroy wellsprings, but not quite there. Hope this sheds some light on the operations at Seatribe!