Brian Swetland on Nostr: There's also still just the standard "simplify and refine a test case to reproduce ...

There's also still just the standard "simplify and refine a test case to reproduce the issue" -- what's the smallest chunk of code that can, in isolation, cause the problem to occur? Which would be useful to have in any case. While doing this, one can often narrow down software-related issues... is some utility function, generated code, interrupt routine, etc, causing extraneous writes under some circumstances, perhaps?