Slashdot Mirror
Can People Really Program 80+ Hours a Week?
ibn_khaldun asks: "A question in light of the EA controversy. I'm an academic researcher who does his own programming -- I have to eat what I kill. In my 35 years of coding experience, any time I try to work on a complex program for more than, say, 60 hours a week (coding, not just showing up) for a couple weeks at a time, I'm just asking for trouble: I generate buggy code and debugging it only makes it buggier. Numerous studies in other fields (law firms, hospitals) have shown that mistakes rise exponentially after anyone works about 50 hours per week (don't think about this if you go to the emergency room at 3 a.m.)." Are these rational working conditions? (More below.)
"Does EA sprinkle magic pixie dust on their serfs to get around this problem, or is the work so trivial that it can be done while pathologically sleep deprived, or are the PHB's so technically challenged they don't realize what is going on? This whole 'death march' mentality seems absolutely crazy to me as a programmer, but appears to be common. Honestly, can someone enlighten me as to how these 80+ hour weeks ever accomplish anything?"
I called in Well today because I went in to work on Saturday. I never would have considered that at my old job but I'm finally starting to realise that if I restrict myself to 40 hour weeks I get a lot more done and I have more time to take care of important things like household chores and family stuff.
Yes, it's possible to work for 80 hrs a week, or even more. However, you need to define "work" here. Would you define reading/replying to emails as work? Would you define attending meetings as work? Believe me, these small numbers add up to take away a sizeable portion of your workday and your energy.
/. as time wasted. It's also an integral part of the process of coding. Hence, i do claim that one can "work" for 80 arse a weak.
If you're talking about pure coding, IMHO, coding is not a tap that can be turned on and off at will. One needs to "get into the flow" to get some real coding done, and this doesn't happen easily (at least for me). I'll often spend hours tinkering around with stuff, browsing some site, but won't for the death of me, be able to finish writing a simple class or stored procedure. Maybe, i'll keep getting stuck, maybe i'll be too distracted, maybe i'll decide to read some documentation instead. Then, suddenly, everything will start happening smoothly and i'll complete a day's work in a couple of hours.
I don't know if it works the same way for others, but i really feel that one cannot just keep coding continuously all day. At the same time, i will also not consider the interstitial time spent tinkering around and writing comments in
This is absolutely true. The lead programmer on my current project works at least 60 hours every week (and has for years) and more than that about half of the time. He's in at 6:30am and usually leaves around 6:30 or 7:00pm and he NEVER takes lunch breaks. Towards the end of the week, any problem that he "solves" quickly usually requires at least a day or more to re-fix later on. Unfortunately, his seniority makes him almost untouchable when reporting problems like this to senior management who see him as "dedicated and just as productive as everyone else". What they don't seem to notice is that he needs to work about 25-30% more hours per week than the rest of the staff just to produce adequate (not great) code.
Does anyone have any recommendations on how to present something like this to management in a convincing manner?
I remember one major coding crunch at the small software company I used to work for: We were preparing a significant set of enhancements to our core product, while at the same time working on bringing to market a major new product. The two projects shared giant chunks of code and logic, so it wasn't completely off the wall, but it was still clear that we were going to have an awfully long, hard slog ahead of us.
As a kind of non-overtime overtime incentive pay, management set up a deal to pay a bonus at the end of the project, based on hours worked above a certain point, with all kinds of complicated sliding averages and whatnot. My office-mate and I crunched the numbers, and realized that in order to get any appreciable bonus at the end of the project, we would basically have to commit to 60 hour weeks for the indefinite future.
Well, you know, neither of us were exactly in our twenties any more. I had a wife and a brand new house and a 45 minute each way (non rush hour) commute; and while I still felt spry and nimble, I no longer felt immortal and god-like, even with the help of Mountain Dew and m&ms. I was in my mid-to-upper thirties. I decided that, while I was still capable of working arbitrarily many hours in a week for short bursts during an emergency, there was no way that my health would stand up to 55 to 60 hour work weeks every week, indefinitely. Both my office-mate and I decided not to bother signing up for the bonus program.
<irony> (A couple of years later, I fell ill with a chronic and incurable medical condition which has left me essentially unable to perform any work at all; so I suppose I needn't have bothered being so careful).</irony>
In fact, much to management's chagrin, only one member of our small programming staff -- call him "X" -- actually decided to commit to their schedule.
Determined to get a decent bonus for his troubles, X threw himself into it, working 60 and 65 hour weeks. In the meantime, my office-mate and I upped our hours, too, but to a lesser extent: 55 hours one week, 52 the next; and so forth.
The weeks wore on, and we inched along towards our various goals. X was doing his usual fine work, but he was looking more and more haggard (we were all a bit worse for wear, actually). His code got a little sloppier at times.
And then one morning, he committed a bunch of working code to the wrong place, and instantly wiped out about 20% of our company's source code repository.
Did we have backups? Yes, we had backups; but still, it took three or four of us much of the day to both restore everything and to verify that everything was correct. The final tally was, roughly, at least one full man-day flushed down the drain in fifteen seconds due to nothing more than pure exhaustion.
Eventually, the crush passed, of course. It is probably a coincidence that X left the company shortly thereafter, although he came back a year later or so. He's a very good programmer, but some of the code he wrote during that crunch -- especially later on -- was, shall we say, sub-optimal.
We were the company that coined the term Internet Time. How did we do it, by sleeping at the office and working close to 120 hours a week. Was it healthy? No. Was it smart? Probably not. Did we produce a good product? You tell me. We wrote Netscape Navigator 1.0 in less than 6 months time. (Please don't confuse it with Navigator 3 & 4, which was a very different team) But there was a catch. We had all written a browser before. We were not trying to dream up a new product completely from scratch. We had a good idea from the start of what we wanted to build. Those set of circumstances don't happen very often. If I was tasked with building a new product that had never been attempted before, I would never try and work that many hours. Good design does not coexist well with exhaustion. There are plenty of other reasons not to work crazy hours as well, one of them is "having a life"... :lou
Oddly enough, I went throgh a period where I found coding after a couple of drinks or joints actually helped produce better code.
It was during a period where (modesty aside) I was maturing from "someone who could program well"[1] to a "good programmer"[2]. I've always been unusually aware of my own thought processes (as I suspect many programmers/hackers/martial artists/meditators/etc are), and I've noticed that good programmers all seem to go through a stage where they stop programming with their left-brain[3] and start more right-brain-thinking[4].
During this time I discovered that, as long as I already understood the problem fully, a couple of drinks (or joints) seemed to help me internalise the "rules" of a language, and spend more time on the actual creative side of programming - solving tasks without spending the whole time thinking about syntax or grammar.
Of course, some of the code was still pretty squirrely (what a wonderful word), but I do remember on several occasions waking up in the morning, running over my code again to check for bugs, and actually being blown away by how elegant some bits were - I hadn't thought I was capable of writing code like that at that point in my education. I remember one time finding a solution to a problem in linear time that I hadn't even realised sober could be done in less than exponential time, and it quite freaked me out for a while afterwards.
Even now (several years and several languages afterwards), I find coming back to a problem after a drink or toke can sometimes help you see "alternative" ways of solving it, often wildly different to how you'd normally go about it...
Fotonotes:Everything in moderation, including moderation itself