I've seen this a lot on the systemd GitHub Issues. A bug is reported, Pottering describes it as not a bug and closes it himself, then everyone else disagrees and says that it needs to be fixed.

Why is this on HN?