How to Pick Your Battles on a Software Team

In the next month we are going to launch a new product, I'm full stack dev and I'm bit frustrated with my colleagues. I did all the devops and all developers(5) has his vagrant per branch, each vagrant box is deployed in one box so they can show to another colleague without trouble (I built a server with the subdomain with the $TASK-ID.subdomain) and the workflow is correct for all of them and they are happy. To deploy to staging/prod we have a full terraform + consul so we're in a happy place.

In terms of code this is painful, they didn't make any test, I made around 50 test 1 month ago to show to them how to make test, but in one month, only one test was added in the project. In terms of the code-review they complain about good indention, and each time that they made a Pull-Request (pep8 and pyflakes) is not passed, and they use to complain about it. They use to mark task as resolved before code got merged, things to weird that I can't explain.

I tried to explain why things need to change, I made a workflow in plantuml and they are complaining that it's too difficult. I'm the only one that made commits in the docs folder.

Nowadays I'm thinking to switch to another company, but I'm happy with my manager (He is too honest), but the team is killing me, and I don't feel comfortable with them, I rarely learn from them and they don't want to do things better, they only want to work to raise money.

This is a dead battle? Or can I fix this?

(1) let go of your frustration, it is not helping you, if you can't GOTO 3.

(2) pick the best of your colleagues and try to lift him/her up to your level with arguments, not with force (this will take time and oceans of patience)

(3) if that does not work, start looking for other employment, no matter how much you like your boss

(4) if it does work, then now you have a buddy, who already knows your method works, rinse and repeat at (2) until you've either had enough of it or you find that you now have a well oiled team.

Good luck! (You'll need it, this is not an easy thing to do). And if you manage to make this work you are management material yourself.

If you're dealing with people, then you are going to have problems with ego. Sure, it'd be great if everyone could lay down rational arguments, then come to a rational decision and rationally agree to move forward. It'd also be great if the unicorn færies showed up to take us all to Neverland. People have emotions; they hold grudges; they feel that fairness demands that if they yielded last time then someone else should yield next time; they have genuine differences in opinion and perspective.

A lot of software is art: there's no mathematically correct answer, just æsthetics or style or simply choice. People will disagree about which choice to take; they will feel ignored if they don't receive validation from time to time.

Now, an individual can choose to swallow his ego, but it's not pleasant — and frankly sometimes it's bad for the product to be quiet and let poor decisions move forward.

So yes, you have to pick your battles.

Please, we should not be so dismissive. "Great team" commonly refers to output, not internal organization. (At least if we use a charitable interpretation. And if we don't, we're in the uncomfortable position of demoting many teams from "great" once we learn of their internal dynamics.) Many teams even often seem unaware of their dysfunctions.

Even if we don't use the charitable interpretation, then perhaps a "great team" would act on the assumption that it is almost certainly dysfunctional, as part of ongoing processes of self-improvement. After all, once you think you're perfect, you're in trouble. (http://model-view-culture.myshopify.com/products/your-startu...)

And teams aren't "closed". Outsiders (like that contractor or executive from another department) interact with the team and are momentarily part of it. Interactions with them may be dysfunctional, due to culture clashes, inevitable bumps as people learn to cooperate, etc.