DEV Community

Discussion on: Politeness or Bluntness in Code Review? Settling the Matter Once and for All

Collapse
 
johncip profile image
jmc • Edited

Different Cultures Land in Really, Really Different Places on the Subtlety vs. Directness Spectrum

I agree. In particular I think that supplicating-by-default can be very problematic when folks aren't expecting it.

"nice or blunt" is ... staking a claim to your power or lack thereof.

Ideally, among knowledge workers, "power" (though I prefer "standing") follows from correctness. If that doesn't happen during code review, the only good option is to leave for a place where it does. Maybe it's naive, but I think a focus on correctness is how you break out of the politics.

At best bluntness should be tempered by:

  • how sure you are
  • how important the thing is
  • assuming the other person has good intentions
  • knowing that learning a thing before someone does doesn't make you smart

I try to be right and nice, in that order. And I try to know when I'm not sure, or when it's a matter of taste.

Folks complain about the tenor of Stack Overflow, but the only time I see people get chewed out there is when they're very confidently wrong. IMO such reprisals are useful for maintaining a culture of accuracy.

Linus Torvalds

Unfortunately no one tells stories about when someone isn't a dick. We've got a big pantheon and I'd argue that most of its members are fairly unassuming. I've never read about Kay, or Steele, or Hickey telling someone they're "so stupid it's a wonder they figured out how to breastfeed." (That's the cleaned up version.)

the right strategy is one that you tune according to your audience.

That was my first thought. If you know the person, use the style of communication they prefer. But be nice. And be right. Ideally, explain why you think you're right, and have the integrity to also say why you could be wrong.

Collapse
 
daedtech profile image
Erik Dietrich

I'm trying to resist the urge to write a novel on the subject of office politics, since it's both something I've talked a lot about on my blog for years, and, in its current corporate incarnation, the thing that eventually drove me out of employment and into working for myself. At any rate, I'd say the fatal flaw experienced by most groups aiming for "correctness as currency" would arise from issues where correctness is either unknowable or else non-existent due to subjectivity.

So, at any rate, I don't take what you're saying as naive at all, but rather the seed of a good mode for interaction. Because if correct is (currently) unknowable, that should result in extremely productive discussions of "this call seems subjective, so how could we run an experiment and measure it?"

For me, that approach tends to eliminate a lot of discord, in the polite vs. blunt arena, but in terms of human interaction in general. "Is this a good blog post" is a question that can quickly become a terminally stupid shouting match between respondents. But, what if you say, "good as measured by what -- time on the page, reader engagement, etc" you can then start to establish outcome-oriented metrics and have more productive conversations.