To not get this wrong: I am a huge supporter of OSF / COS and think they did an incredible job to make science more open and transparent to everyone. Stuff like this can happen, it’s good that they are identified and reported. Let’s just hope that incidents like this make the platform even stronger.

A recent redesign of OSF by @cos.io led to widespread access failures. What began as a few broken download links became for me a total disappearance of eight years of DOI-registered work. What happened, how was it resolved, and what it reveals about trust and infrastructure in open science