This is something I’ve been wondering about for a long time. Programming is an activity that makes you face your own fallibility all the time. You write some code, compile it or run it, and then 80% of the time, it doesn’t work exactly the way you imagined. There’s an error message, or it just behaves incorrectly. Then you need to iterate on it and fix the issues until you get the desired result, and even then it’s subtly wrong, and causes an outage at 3am on Sunday.
I thought this experience would teach programmers to be the humblest people in the world.
I can’t believe how wrong I was. Programmers can be the most arrogant dickheads you will ever meet. Why is that?
TL;DR; Human emotions don’t belong in software development. Code doesn’t care about your feelings so neither do senior devs. If you want someone to listen about your feelings, talk to your manager, not other devs while proposing a solution based on feelings. If you can’t handle critical criticism of your code/designs/etc, you simply don’t belong. If you are pumping shit out that takes down prod at 3am on Sunday, it is exactly that… shit.
It’s a problem if you’re a junior dev. Junior dev acting like they know everything straight out of college will often just get ignored and pushed sideways in the company (rather than promoting upward).
Junior devs should be properly mentored by a senior dev. Junior devs should never be placed in a position where their code could “cause an outage at 3am on Sunday”. If that happens from a junior devs code, that is mentor’s fault all the way.
Intermediate devs have more leeway since they have seen some shit and often decide and design based on real industry experience rather than a couple of college courses. However, intermediate devs are often deeply passionate about their projects, especially when they are doing something like “designing specifically so we won’t have downtime at 3am on Sunday again because that sucked”.
Most humans have an extreme difficult time separating emotion and logic. When intermediate dev above “knows this is right”, and passionately debates the topic to no end, even when there is an obvious flaw in the logic, they are often debating based on their feelings and not based on matters of fact. This is a problem, and will come off as arrogance, but most likely is not actual arrogance and they are just caught up in their feelings about how hard they worked on the code/solution and they “feel” (not know) that the solution is correct/best/etc. This is easier than saying “damn, I’m wrong” for most people.
Senior+ devs often have seen shit across multiple softwares and companies allowing for more generic context surrounding issues. Senior+ devs also often depend on who they are conversing with, and often change tact to the situation. Ignore junior devs unless it’s their actual job (mentorship). Be a contrarian to intermediate devs, to ensure they are doing real due diligence with their solutions. Heavily debate senior+ devs, as this is often where emergent architecture and design comes from, and all parties come out more knowledgeable. Caveat being: debates must be about the data and facts, with no ad hominem.
There is a major difference in the following statements:
The former has nothing to do with you, even if you wrote the code. The code would still be dumb if someone else, even me, wrote it. Too many people take criticism of code as personal attacks – human emotions do not belong in software development, facts and logic do.
I’ve seen variations of this opinion a lot in tech, but you don’t have to sacrifice frank feedback to be a little more polite. Like sure “this code is dumb” is better than “you are dumb”, but not as good as “this could be more efficient if we did y instead”
The latter is more helpful and less confrontational. The best devs I’ve worked with have been excited about code, and that came through in review. When they saw something that was wrong their feedback came across as excited to share rather than affronted by ignorance.
The thing devs don’t like to admit is that a ton of stuff we argue about is more taste than science. There’s no right answer to the classic performance vs readability thing, or where exactly the line is for effort vs payoff. Couching feedback around things like that a opinion often makes people way more receptive to it. “Hey I’d prefer if we did it this way” often goes a long way, IMO.
I actually welcome constructive criticism, and the brand of arrogance I’m so frustrated with is when people dismiss ideas or arguments without offering valid counterarguments. Or maybe when they take the worst possible interpretation of what someone says (see example above) and argue against that. Maybe my original post wasn’t clear enough.
While it’s true that separating emotions from the work itself and learning to accept justified criticism is important, even crucial, this fact doesn’t give free license to the people giving the criticism to be rude. Your comment seems to imply that what I perceive as arrogance is often justified when “senior devs” are defending their solutions based on their experience. But arrogance != conviction or confidence. Confidence paired with humility allows for open-mindedness, and creates a better environment for everyone involved. It encourages sharing and discussing ideas.
You also state that human emotions don’t belong in software development, then proceed to write a long rant that reeks of condescension, dismissiveness, annoyance, frustration and a feeling of superiority. Your comment is by no means neutral and analytical, it just displays emotions you feel are justified, while those felt by others are not. It’s hard not to conclude that you are exactly the type of person my original post is about.
Damn, I was wrong… ;)
I guess I should also clarify (if it wasn’t obvious), I am speaking about enterprise software development at large scale, and probably none of this properly applies to single OSS with 4 devs (or similar).
My intent was not “senior devs ok to be arrogant”, but more so give my understanding of why a lot of senior devs come off as arrogant (if code is dumb, it is okay to say it is dumb, and is not a personal attack).
Simply saying “this code is dumb” and walking away is arrogant – agreed.
Commenting in code review “this code is dumb” and nothing else is arrogant – agreed.
Saying “this code is dumb” and debating the concepts at play in code, is not arrogance, but is often perceived as ad hominem (“well [dev] invested tons of time and energy on writing that code you just called dumb and is offended” – why is [dev] working in a silo to the point that the code is trashed in code review on pull request?).
Correct, I was trying to give a why, directly from someone often perceived as an arrogant asshole senior dev.
Maybe it is better to say “feelings belong in process, not code”. I consider everything I discussed process. Code taking down prod at 3am is not, but more of an indicator of bad process (in whatever form).
This is interesting. My comment doesn’t reflect my actual feelings on this topic, but rather my current understanding of this phenomenon I (clearly) struggle with, based on my experience in industry.
My actual feelings align more with “I am blunt and don’t care about your feelings. This is a business not a personal project, get over it.” and acknowledge that is arrogant, shitty, and not helpful (probably the other fancy words you used too) but saves ME hours upon hours weekly, which I then get to work on my code/tickets during, which is what my job actually is, and is the ACTUAL metric by which success is tracked.
If you are looking for design review, ask an architect (not through PR to main, which most often indicates “ready”), that is their job, not developers. If you ask for my review I will give it, and that is the “to be nice part”. If I look and want to comment “this code is dumb – rant”, and I am the proper reviewer, I ask my manager to help me navigate the best way to express myself without coming off as an asshole.
I also acknowledge I do not write the checks, and have many times over written “dumb code” when that is what the business wants. Although I do make sure and be clear that it is dumb code (business people dgaf until the downtime happens tho).
I think what you are saying is technically correct, but it fails to acknowledge the importance of human emotions. Emotions are involved in everything humans do, and since software is written by humans for now, the act of developing software also involves emotions.
This does not mean that it should be a emotion-driven discipline or that emotional arguments should be weighted more than other kind of arguments. It just means that everyone will deal with human emotions while developing software, both their own and others’. We can of course manage our own emotions as we wish, but then the question is how do we act on regard of others’.
My main thought here is that disregarding how others feel about your actions rarely helps anyone. When collaborating with others, we usually aim for getting the most out of that collaboration. That is only achievable if the other person feels safe and welcome. Maybe writing some harsh comments in a code review is not the best way to do that.
Next thing that needs consideration is that we are all different, and what for one is a small gesture, for other might be offensive. You don’t know everyone else’s personal history, background or experiences, so you don’t know what can negatively affect them. This needs to be acknowledged, particularly when working in heterogeneous groups (ie, people from other cultures). You can ignore this fact, but it will negatively affect the collaboration.
As with everything else, people can become better at communication and collaborating with others, and still be able to defend your points without being aggresive.
Why don’t they, then? My first guess would be that this is seen as irrelevant. This is how I do things, how we do things in this industry, learn to deal with it. I don’t think that is a good approach. It does not harm your engineering skills nor the product of your work to be kind to others. On the opposite, it makes their lifes easier. I don’t see a reason why we shouldn’t strive to do that.
I totally disagree with a lot of this and encourage you to rethink how you communicate with peers.
Calling something someone made dumb is mean. Believe me, I totally get where you’re coming from and wish I could separate the emotional aspect from it but it’s just not possible. Humans are emotional beings. We make decisions based on emotions and use logic to explain them. You can’t just say something that someone made is dumb and expect it to not hurt their feelings because it “would still be dumb if someone else, even [you], wrote it.”
It is possible to critique code respectfully without losing anything. I realize you’re likely simplifying for the sake of discussion, but consider these examples, which would you rather be told?
These both convey the same message (the code is wrong and how to solve it) but one is polite. The first one is in the gray area and I choose it for that reason. It may very well be true that it is a mistake that amateurs make but saying someone made and “amateur mistake” can be degrading.
People view rude words about their works as rude works about themselves and any appeal to logic will not change that. Just say what you would say but be more polite about it.