text stringlengths 0 2.35k |
|---|
**Anderson Queiroz:** You have to comment on it. So it's really hard to comment on it, at least for the tools that I use at GitHub and the IDE. It's hard to comment on the code. If it's something that for some reason -- you know when you do either a greenfield project, or the pull request is a huge refactor... You know... |
I've done it once or twice, and then I created my branch out of this branch, and then I open up pull requests for this branch, so the person can see my comments in the code, without having to go look for it. |
But this was pretty much in either greenfield projects - you know, when you start something new, or when you're just adding so much new code that the pull request itself, it's hard because it's completely out of order and everything, and there is a lot to comment on. |
**Natalie Pistunovich:** Yeah, that's the hardest ones, when there's so much to handle there. |
**Anderson Queiroz:** Yeah. |
**Natalie Pistunovich:** Do you sometimes find yourself re-reading the whole thing to kind of - once to read it, and second time to make sense? |
**Anderson Queiroz:** Yeah, definitely. Going back to an old Go Time, when you talked about documentation and reading documentation - I have the perseverance; if I'm not understanding, I'm gonna read it over and over and over again, and I'm gonna do my best to understand. If I don't, I'm gonna ask. But I think if somet... |
**Natalie Pistunovich:** Yeah. |
**Angelica Hill:** Do you comment as you go, or do you read through, fully digest, and then go through and do all your comments? |
**Anderson Queiroz:** No, no, I comment as I go, as a string, and then when I get down there, I'm like "Oh yeah, that's why!" So I go back and delete the comment. |
**Angelica Hill:** Oh, okay. |
**Anderson Queiroz:** You know, you're like "Dude, it doesn't make sense." Something like "Why?! How does it work?" And then you read it, "Ahh...!" And then I go back, delete and edit the comments. |
**Angelica Hill:** You go back, "Oh, I'm so sorry. Never mind..." |
**Anderson Queiroz:** I mean, on GitHub you don't submit the review, right? |
**Angelica Hill:** Yeah. |
**Anderson Queiroz:** So I just delete. But the feeling is exactly that I'm apologizing for asking something stupid. \[laughter\] |
**Angelica Hill:** When you're going over a PR, do you feel like - or any reviewer - should test the changes? And to what extent should you test the changes, if you think that they should be tested by the reviewed? |
**Anderson Queiroz:** I think that's the one million dollar question. |
**Angelica Hill:** Give us the answer, we're ready! |
**Natalie Pistunovich:** Maybe another unpopular opinion... |
**Anderson Queiroz:** No, I think the answer is it depends... |
**Angelica Hill:** Okay. |
**Anderson Queiroz:** We say that a lot at Elastic. But I think it depends on your role on this pull request. Imagine that I'm a maintainer of open source, and I'm taking something in from a person that's not from the wrapper, I'd be extra careful. So perhaps I would try to run the code. Usually, I don't really test th... |
**Angelica Hill:** Okay. And does the length of the PR or the scope of the change change that opinion? |
**Anderson Queiroz:** \[30:06\] Yeah... |
**Angelica Hill:** Okay. |
**Anderson Queiroz:** I think the length not so much... Let's be honest, the longer the PR, the less detailed is the review. We are humans, we get tired. Come on. If you're reviewing like 15-20 files, the last one - you're tired. It's just a human thing. And it's something that I've done a few times. I don't like it, b... |
**Angelica Hill:** So given that, what is a reasonable time to expect a PR review? If you put in a PR today, is it the next hour, by end of day, a week? Does it depend? |
**Anderson Queiroz:** The real answer is "It depends." |
**Angelica Hill:** \[unintelligible 00:31:00.20\] |
**Anderson Queiroz:** Yeah. At my first job we -- did we use JIRA? I don't know, whatever; we used columns. We used columns, so there was the column "PR reviews." So the open PRs were there for review. And then there could be -- I think we were three people, so there could be only two PRs on review. So do you wanna put... |
**Angelica Hill:** Okay. And do you have different commitments when it comes to your internal team PRs, versus people who are maybe contributing to your service? i.e. like in our system we have a lot of external teams that will contribute to our service and ask for PR reviews... What is a reasonable timeline to commit ... |
**Anderson Queiroz:** I think there are two categories. If it's just a normal flow, they can go in the same flow. But if it's something that someone external is doing, because our team doesn't have the capacity, and then it is really important, probably I will try to prioritize this review. |
But also, if someone that's not from the team, who doesn't know the wrapper conventions, it's probably gonna be a more thorough review. I strongly believe that at least your code should be consistent. I'd rather have something that I don't like, but it's consistent and it's always there, than half of the code I like, h... |
So in extended reviews I think there's the extra consistency in code conventions from the wrapper that you have to put through. And then it's better to be quicker. |
**Angelica Hill:** For sure. I really like that column policy. I might have to implement that on my team. |
**Anderson Queiroz:** Yeah, right?! |
**Angelica Hill:** You can't put a PRon if there are already two. You have to review them. I love that. |
**Anderson Queiroz:** Yeah, it makes things to move. And I think it's nice. |
**Angelica Hill:** I agree. |
**Natalie Pistunovich:** It's a waitgroup. We've pretty much described this concept now. |
**Anderson Queiroz:** It's a channel with a buffer. |
**Natalie Pistunovich:** \[laughs\] Yeah. |
**Angelica Hill:** And then I'm gonna have to be the bad girl who comes into Slack and someone's like "Oh, I'm ready to put my PR in this big new feature and I'm like "You're not allowed to..." |
**Natalie Pistunovich:** \[unintelligible 00:33:24.15\] |
**Angelica Hill:** "Go review Bob's PR." |
**Natalie Pistunovich:** Throwing error... |
**Angelica Hill:** \[laughs\] |
**Anderson Queiroz:** Yeah... There's like the poking PR review. You only get your PR reviewed when you poke someone. You don't want that. |
**Natalie Pistunovich:** Exception. \[laughs\] It could be a fun way of teaching all sorts of Go concepts, now that -- this gave me some ideas. Thank you. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.