Joël Stemmer

I've had a quick look at the fastWalk implementation and it looks like it should be pretty easy to replace the current filepath.Walk implementation with this faster one. Some quick...

Sure, I'll hold off on sending it for now and will keep it around in a local branch until gps is merged into dep. I was vaguely aware that symlinks...

Are you running Jenkins on a Windows server, or is this something specific to Jenkins itself?

Thanks. It sounds like we should at least strip the invalid characters from the CDATA section. However, since ANSI escape sequences also contain regular characters we'd probably want to strip...

@janisz Where do these invalid bytes come from, this doesn't look like an ansi escape character. Any concerns if they are just dropped completely from the output?

Leaving this issue open for now, because it would be nice if we could also detect ANSI escape sequences and remove them from the output.

I haven't used the fuzzing yet in Go 1.18, but a quick test shows the output is somewhat similar so it might work already. Have you tried it? If you...

Thanks, that's very helpful. It looks like we should be able to support this with a few small changes, and in fact it looks like the v2 branch already outputs...

Thanks. It looks like the testcase gets attributed to the wrong package, possibly because of the order in which it appears in the output. I don't think I've seen this...

I managed to reproduce by testing multiple packages at a time and enabling the race detector (`-race`). For non-JSON output, test events are always grouped by package. When enabling JSON...