Comment by ramirez60 | Hacker News Reader

Comment by ramirez60 | Hacker News Reader

Parent Full thread

I don't mean to claim doctors do fine without enough sleep. Doctor friends I have spoken to talk about this talk about the reforms but it is not uniformly positive. I am trying to posit that the claim that we function better with more sleep is true. But that claim is made in a vacuum. Person A operates at 100% well rested. But operating at 80% or 60% is more valuable than 0% if averaged out over time. So I can write better code when I'm well rested, sure. I agree with that. I think everyone logically will. If I have 1000 lines of code to write. And it will take me 10 hours to write it. And I have 2 days to do it. I'll spread it out over 2 days and write good code, and maybe it only end up taking 800 lines of code. But if I have 1 day to write it. I might end up writing 1200 lines of code, but still get it done in that day, it is more valuable to me, than a day later when the code doesn't matter anymore. I'd argue I'd rather a doctor save my life that day, than do a great job and make sure I never scar over two days when the option isn't there.

Medicine is a regulated industry in that the number of people trained almost precisely matches the number of jobs available. The training process is designed to prune out people who cannot function when they have to choose between life saving procedure, and doing the life saving procedure perfectly but losing the patient. It's a straw-man I know, I'm just tired of everyone being absolutely fascinated by this idea that resting is better than not resting. Isn't that just obvious? Does anyone feel like they work better when they are tired? What knowledge are we gaining by these discussions?

If you are throwing away your tired-brain code a week later, godspeed. Do what you like. But if you are checking in that code and making other people deal with it, then quality becomes much more important. If you write 1200 lines of code instead of 800, then you have something that has 50% more maintenance cost and 50% more bugs. (And it's probably worse given that you were tired and thinking with a short-term mindset.) So hell yes, it matters to me whether my coworkers are putting long-term value above short-term incentives.

> I'm just tired of everyone being absolutely fascinated by this idea that resting is better than not resting. Isn't that just obvious?

It is not in fact obvious. So many places reward time spent more than quality work. Just today at a meetup we were talking about how to shift people away from damaging and expensive heroics to investing in not creating such big problems that you need work-all-night heroes. The industry is notorious for death marches and work-all-the-time startups. And it's also notorious for buggy, low-productivity, high-tech-debt code bases and giant rewrites caused by years of accumulated short-term thinking.

If it's obvious to you, skip over the articles. Not everything has to be for you.

> I'd argue I'd rather a doctor save my life that day, than do a great job and make sure I never scar over two days when the option isn't there.

> What knowledge are we gaining by these discussions?

Sure, if that was all there was to it, then you'd be 100% right. The knowledge we seem to be gaining from these discussions is that a programmer who works 6 hours a day is overall more productive than one who works 8 hours a day, let alone one who has been on a 16 hour death march for 6 months. The same idea applied to the doctor example may mean that your doctor is less likely to save your life if they're on the 28th hour of their shift than one that wasn't.

Yes, there's a time and a place for burning a candle at both ends, but we seem to think that this is OK to do in the long term. So sure, if you need to pull a rare 20-hour marathon to fix an urgent problem, do it - the value might well justify the cost. But don't tell yourself that it's sustainable or even optimal in the long term, evidence seems to suggest the exact opposite.

Reply on news.ycombinator.com