You know how every time you point out a problem to legacy publishers — like when they’re caught misrepresenting their open-access offerings they explain that it’s very complicated and will take months to fix? Here’s how that should work: To summarise: I found a bug in the PeerJ system;