Comment by gorgoiler | Hacker News Reader

Comment by gorgoiler | Hacker News Reader

The hardest thing about remote work is code review.

What used to be a collaborative effort where we both take responsibility for making things better now feels confrontational because online code review is now often the only interaction we have with others.

It’s crucial to build trust and alignment with your colleagues. Otherwise you end up blaming all your problems — which, in fields like software engineering and site reliability, shower down upon us with the misty subtlety of a firehose — on the people who you only know by their Unix name / commit email / cat themed github avatar.

The die hard “work to live not live to work!” responses, here, don’t seem to be addressing that. It’s a much deeper problem than whether you should make friends in a bar versus by the water cooler.

Working in a team is social and you have to form social bonds. No one is saying they are the only social bonds you should form in your life. If you refuse to form any social bonds at work then you will not be as effective as someone who does.

Code review has been done "remotely" via Github or other online tool at every place I have worked, I have _never_ seen it done in-person. I also find that doing this in person would be incredibly time consuming and frustrating. The confrontational nature of code review is a good thing because people are able to focus and independently analyze how your code works. Its important to be nice but its also important to not take feedback about your code personally.

Thanks for your comment. I amended my second paragraph to try to make my point clearer.

I am in no way advocating for in-person code review. I’m saying that it’s much harder to do code review with someone you haven’t shared lunch with, chatted with in the pub, or had a non-work related phone call with for 9 months.

Your point about public GitHub is pretty much on point. Some of the least productive and most barbed back-and-forth you see on code reviews is the public FOSS ones where people don’t really know or trust each other that much.

Some of the best are where the reviewers clearly want to work together on something and not be confrontational.

(By the way: I don’t see this behaviour as openly confrontational in terms of deliberate aggression — more commonly it’s a reviewer subconsciously expressing their self-perceived lack of standing by criticising other’s work.)

> The hardest thing about remote work is code review.

code review was done async. and in all practical sense "remote" at all the dozen places I ever worked in an office.

How else do you scale code review on a team of more than 2 people. Get all 5+ people in a room to do code review for a 5 liner?

Thanks for responding. My original point didn’t come across well.

I meant that now we never hang out, we don’t trust each other any more, and code review has suffered as a result.

It’s especially hard for new people joining a team where baseline trust is basically equal to that of Internet stranger.

> I meant that now we never hang out, we don’t trust each other any more, and code review has suffered as a result. > It’s especially hard for new people joining a team where baseline trust is basically equal to that of Internet stranger.

Huh. How's your hiring process like? Do you hire very young people?

The only time I've seen this happen is when the team used to work in a colocated manner, and then some members started to WFH. Eventually some of those people WFH got too tied into the "home" part of WFH and work suffered. Simultaneously other people in the office just started to treat those WFH as second class citizens as well. This really caused a mess.

If this is what you're seeing, the easiest solution is to just move to a remote only team.

If you want to stay, you need to have a good manager in place who understands these dynamics and can mitigate issues.

Reply on news.ycombinator.com