Working Late Makes Me A Bad Developer
I must be an outlier, at least I feel like someone who doesn’t belong in the tech industry. I’ve got a real issue, and it’s caused by one simple fact.
I’m afraid of working late into the night.
The Ghost in the Machine
I’m not afraid of the dark. I don’t think there is a skeleton in the back of my closet. I have concerns that I might run out of coffee or let loose my opinions about JIRA in an inappropriate context, but that’s not the topic of today’s blog post.
No. I’m scared about the impact of late night working sessions on my code. Am I the only one?
A Romantic Vision
I’m not some romantic moonlight hacker building a unicorn startup from my kitchen counter. I’m someone different entirely, come dark I’m a different software developer.
I transform into a tired, panicking software developer who no longer understands what they’re even doing. I might add and remove the same code three or four times before settling upon a (trivial) approach, asking ChatGPT and cursor for advice every step of the way.
The next day when I try to open the pull request and ask “Who wrote this?”, I wonder how such poor code is possible and why I thought it might compile last night (it doesn’t in the morning light). All the blunders and misunderstandings were me, they were always me.
The Issue
It’s not just the quality of the code that gets messed up bad. The whole atmosphere around late-night working causes me issues and fear. I develop a brain fog, and deliver code littered with mistakes revealing my sheer exhaustion.
I might begin the evening thinking that I’ll be able to leave on time. After mid-afternoon the pressure begins to build, will I be able to make progress on this ticket before the end of the day? The pressure isn’t always from a deadline, it might be wholly from me wanting to get things done and prove that I’m productive. The pressure comes from trying (hard) to avoid looking like the weakest member of a team in which I’m certain that I don’t belong.
Yet I know the truth. The later I work, the worse the outcome and the worse developer I look to my colleagues. The cost of fixing late-night mistakes is higher than leaving the work for tomorrow. I’m sure that people are going to love criticising me in the comments for not being hardcore enough for their tastes. Yet there is perfectly good research that tells the same story, working late tanks memory, decision-making and productivity. Poor sleep costs hundreds of billions in lost productivity each year. You think staying up to fix a bug helps? It probably costs more than it saves.
Scared to Stop
The problem is I can quote studies, and I can tell you that working late isn’t good for anyone yet I still feel compelled to do it. I’m scared to stop.
I don’t know if you’re in the same mental position. The pressure to stay online, to be seen working, to solve everything before logging off even if it means rolling into bed at midnight with a half-broken feature and a half-working brain.
Those colleagues who send Slack messages at 10 p.m. contribute to the feeling that that’s simply “what it takes” to be a success. Maybe I’m completely alone in these dark feelings, but I don’t think so. Yet I’m convinced that feeling alone is part of this.
That’s because the culture of software development is one of late-night glory. That’s a destructive culture but it’s one that we all seem to be part of.
Burnout Glory
The developer who works until 3.a.m. is faited as a 10x engineer burning the midnight oil. It doesn’t seem to matter if you cause a production incident the next day, or whether your hardcore working hours cause the error. The fact is that you burnt the midnight oil intending to fix a problem, and you’re great since you *fixed* the issue, and who cares if it is a fix that involves some rework in the future?
As an industry we don’t talk about the developer who breaks the code because they were too tired to read an API properly. We don’t talk about the anxiety of looking up who else is online in the middle of the night, or the quiet dread of waking up the next morning and realizing you didn’t ship a feature, you shipped a mess.
So here it is. We should be developing (alongside our code) a line in the sand which will we’ve written before 7 p.m.
Inverse heroism
It’s not cool to work on a feature into the early hours of the night.
Late-night work isn’t heroic.
It’s lonely. It’s inefficient.
It’s time we stood up for ourselves and said the unchanging truth. Working late is a failure of planning, not a badge of honor.
That isn’t the same as working because you want to, figuring out a new API or getting a new feature done so you can clock off on time tomorrow.
It isn’t heroism. It’s reverse heroism.
Conclusion
Working late because of what people may (or will, you know they will) think about you isn’t cool. I want you to know. Stopping work and having standards doesn’t make you weak.
Knowing you have limits and enforcing them isn’t a sign of weakness.
It’s a sign of wisdom.
So go to bed. Have a rest. Be intentional in the work you do and the results you’re doing to get.
Go to bed. Your code will thank you.
About The Author
Professional Software Developer “The Secret Developer” can be found on Twitter @TheSDeveloper.
The Secret Developer writes their best code between 9 a.m. and 10 a.m. It’s a pity about the clash with standup, but nevermind.