this is so alien to me. are you and GP talking about putting in long days during crunch time, or actual on-call rotations where you're expected to be available in the middle of the night? i agree the former is universally expected, but the latter? not at all, not without discussing it beforehand. that would be obscene. i am so boggled to hear people accepting and defending the practice.
I kind of locked in on my recent experiences when answering that. So I want to clarify -- its not expected in all orgs and is dependent on roles (e.g. if you aren't working on live services, it wouldn't even make sense). I would expect larger shops to have dedicated ops people or site reliability engineers who handle most of the duties. But I think small to medium sized corps that either don't have or cannot afford these would expect developers to handle crashing services, whenever they occur. Having an on-call schedule is a natural next step to prevent everyone from always being on call, or (worse, imo) prevent people from siloing and only fixing "their" stuff. I agree it should ideally be part of the job description, but don't find it unusual to be left out if the job otherwise fits the mold.
