That's a no from me on every point of this argument. I don't really like logging, and I think some ecosystems do it better than others (nod-cough), but this art

That's a no from me on every point of this argument. I don't really like logging, and I think some ecosystems do it better than others (nod-cough), but this article doesn't actually make anyones lives easier. Logging each exception vs logging nothing is a false dichotomy. I'm not sure anyone whose suggested fudging a 'Result' type in a language without sum types has ever done so, or at least stuck around afterwards. I'd like to be equally condescending in my answer to 'is my team ready for logging side effects?'. The fact that a distributed log aggregation, storage, and visualisation framework has a docker-compose file no more complex than a well configured SQL database does not terrify me.