3 ms·

I would say screw papers, make a web document, make it easy to repeat your experiments, publish code that work as is.