Forward propagation of errors through time (nicolaszucchet.github.io)

Really appreciate how they're publishing failed research. It's much more needed in academia to know why a method or idea doesn't work - which nobody shares since it doesn't help with publishing papers in top journals and earn grants!