I think I’ve blocked out most of the bad ones, and most of my coworkers have been good or okay.
This one guy I remember though. This might only be relevant for people who live in software.
He refused to name his tests. Normally with jest or mocha you have tests like
describe(“user settings page”, () => {
it(“allows the user to change their name”, () => {
etc
But he refused. He’d put empty string in both spots. So you’d open a test file and there’d just be a dozen anonymous unlabeled tests and you’d have to puzzle out what they were trying to do.
He was a reasonably nice person when we talked, at least. But this drove me crazy.
A previous workplace had a former contractor I constantly had to fix his shit from. One of his many peccadilloes was how he did shell scripts. He’d have dozens of scripts in folders, each named things like “bsr001, bsr002, bsr003,” and so on. Each script was only three lines.
The usual #!/bin/sh
One line of code.
A call to execute the next script in sequence
Some of his stuff was “encrypted” by base64 so you couldn’t read them. I mean, it was easy to figure out how to decrypt them, but still annoying.
Guy charged $250k/year for his work. And apparently was so surly, when I started working there after him, they were shocked how friendly I was.
He said that the test names were essentially comments, and “comments quickly become lies”. Which, fine, we’ve all seen bad comments. But the test names are more like file names than comments, and no sensible person is going to suggest we get rid of file and folder names. Except the other guy who responded to me with the shell scripts with no names that each call each other, maybe.
He was on a different team at a large company so I didn’t get wind of this right away. We had a meeting scheduled to hash it out, but then there were mass layoffs that day and I left shortly after. For all I know he’s still there
I had a boss who didn’t allow comments. “I like clean code,” was what he said. He also didn’t like variables with easy to understand names, like config_file_path because he said, “this is a real company, not kindergarten.”
I once worked with a team who would have useful tests like (it’s been years since I used Jasmine so not going to try), but things like if true === true.
I think I’ve blocked out most of the bad ones, and most of my coworkers have been good or okay.
This one guy I remember though. This might only be relevant for people who live in software.
He refused to name his tests. Normally with jest or mocha you have tests like
describe(“user settings page”, () => { it(“allows the user to change their name”, () => { etc
But he refused. He’d put empty string in both spots. So you’d open a test file and there’d just be a dozen anonymous unlabeled tests and you’d have to puzzle out what they were trying to do.
He was a reasonably nice person when we talked, at least. But this drove me crazy.
A previous workplace had a former contractor I constantly had to fix his shit from. One of his many peccadilloes was how he did shell scripts. He’d have dozens of scripts in folders, each named things like “bsr001, bsr002, bsr003,” and so on. Each script was only three lines.
Some of his stuff was “encrypted” by base64 so you couldn’t read them. I mean, it was easy to figure out how to decrypt them, but still annoying.
Guy charged $250k/year for his work. And apparently was so surly, when I started working there after him, they were shocked how friendly I was.
This is some next level awful. Reminds me of that “inner json effect” dailywtf post from ages ago https://thedailywtf.com/articles/the-inner-json-effect
What was his reasoning? Did you not do pairing and or pull requests?
He said that the test names were essentially comments, and “comments quickly become lies”. Which, fine, we’ve all seen bad comments. But the test names are more like file names than comments, and no sensible person is going to suggest we get rid of file and folder names. Except the other guy who responded to me with the shell scripts with no names that each call each other, maybe.
He was on a different team at a large company so I didn’t get wind of this right away. We had a meeting scheduled to hash it out, but then there were mass layoffs that day and I left shortly after. For all I know he’s still there
I had a boss who didn’t allow comments. “I like clean code,” was what he said. He also didn’t like variables with easy to understand names, like config_file_path because he said, “this is a real company, not kindergarten.”
I once worked with a team who would have useful tests like (it’s been years since I used Jasmine so not going to try), but things like if true === true.
Yeah thanks dipshit.
You should have added a linter which blocks PRs with unnamed tests