[maerF0x0]

> low self esteem from ... non-constructive or overly harshly presented critique

ok firstly Comments about the individual in a CR is way out of line.

There is a massive difference between "This code will fail in <a certain way>" and "You failed <, again, like you always do>" ... The latter is out of line and should be addressed head on with a manager present or a person who will enforce it's toxicity.

Now that said, criticism of code should always been allowed. It either has certain properties or it does not. A discussion of economics (scope/timelines/applicability) are a middle ground. But either the code has a race condition or it doesnt. It's easily read (which is best judged by the reader, not the writer)... It's missing some property that is desirable.

If one takes a criticism of code as a criticism of self, that is ego crossing over to where it doesn't belong. I like to remember The bridge will either hold the load or not regardless of how one feels about it. "This wont hold 1 ton" is better information than "Have you considered what happens if you put 1 ton on this?" .. The former informs you of something that must change, the latter leaves uncertainty and obscurity where infact the answer is already known.

[chefandy]

Conversations aren't pure textual exchanges and people aren't robots. Things like subtext, tone, word usage, appropriateness of timing, (like interrupting a presentation,) the volume of your voice, who else is present, the examples that you give and analogies you make to explain yourself, and plenty of other factors weigh into how something lands. If someone wants to conversationally bully someone or make them look unprepared or incompetent in front of people who need to think highly of them, it might not even require criticism. TV journalists do it all the time without making statements at all– they just ask questions.

Being direct and being polite aren't mutually exclusive, and direct criticism doesn't have to leave out any information at all. Saying "This approach won't work under peak load because" in a code review conveys the same exact message as "Your code will collapse like a popsicle stick bridge the second we hit peak load because" during a meeting.

[maerF0x0]

Part of the issue is that "Polite" in american has trended towards indirection. "I'm not sure that was the best choice" is American for "That absolutely was not the best choice and must be reverted immediately" .

[chefandy]

A) That's not an American thing. US confrontation etiquette is still far more direct than British, for example, even if less direct than in, say, scandanavian countries. B) If the message ends up being the same, what is the actual negative effect? I'm not saying you have to like it, but it seems to be an entirely aesthetic "get off my lawn" sort of complaint. Giving shitty, incomplete feedback is just as possible using direct language. C) Since we're talking anecdotally, the people I've met that gave great, frank, direct feedback just did it because they out obligated to, and even in art school with young students, I've seen few people get defensive. It was calm, well-reasoned, and completely no bullshit. Then there are people who might have said very little, or maybe even a lot, but did it in a shitty arrogant tone and that got everybody's hackles up. Everyone I've met aggrieved about not feeling free to give direct feedback was the second type. C) Directness is just a slice of a much larger topic.