tag
9 posts
A safe workflow for post-update Haiku app failures using Terminal evidence, package ownership, prior states, settings isolation, and reports.
How to verify FreeBSD dump capture, preserve vmcore evidence, match kernel symbols, read crashinfo and kgdb traces, and report a panic.
A production-safe FreeBSD DTrace workflow for loading providers, narrowing probes, aggregating syscalls and stacks, and controlling tracing overhead.
How to preserve Haiku crash evidence, inspect the faulting thread and stack, distinguish symbols from guesses, and file a reproducible report.
The difference between an oops and a full panic, how to read an oops call trace to identify the responsible driver, and when it's safe to ignore.
Why a zombie process is actually harmless dead weight rather than a resource leak, and how to find the parent process that's actually responsible for clearing it.
The two essential shell debugging tools: one traces exactly what a script actually executes, the other catches whole categories of bugs before it ever runs.
What the preserved 1947 Mark II logbook proves about its famous moth, what later retellings invented, and what Hopper actually achieved.
Using strace to see exactly which system calls a misbehaving program makes — often the fastest way to diagnose a problem with no useful logs.