Closed appurist closed 3 years ago
The PR was one logical (associated) set of changes. I had already explained privately why it was three separate commits.
The project is not set up for testing this functionality in automated tests. So I manually ensured full coverage by forcing each condition in the debugger, single stepping through the changes. In the future, I'd also like to see a SonarQube report on code quality differences on every commit. but I doubt I'll see that.
There's a big difference between a project with full-time developers and one with ad-hoc occasional volunteers providing fixes, especially where that is fixes to an existing code base with a decade and a half of grunge and code smells. Be careful setting restrictions on what fixes you'll accept, especially when you are the one asking those developers to help provide a fix. Especially if the restrictions are impractical or even unreasonable, such as suggesting commits that are making related, dependent changes be provided as separate PRs.
Several fixes while I was in there looking at Linden plants:
llGetObjectDetails
calls withOBJECT_PATHFINDING_TYPE
to returnOPT_OTHER
for Linden plants.