dybr

Still on pager duty, but nothing happens. Cool.

So I was working.

Last week I could not figure out what happened to the API I deployed to staging. Asked Qian. He had no idea.

So today I investigated. And eventually I figured out: it was not deployed.

Started looking into that. Kaboom, a weird error from kubernetes. And that's why? Found the logs eventually. Wow, the app just does not start. NPE.

And then I figured. It was a pretty bad macros: if you specify a type of a value, <code>s/val x = something/val x: SomeType = something/</code>, the app does not compile in IntelliJ, but it does compile in sbt (and in sbt in jenkins), but throws an NPE. Але, Женя! Oh, Eugene is now busy with other weird things at Google.

So, I wrote a one-liner PR to fix it, and bumped int Chris' resistance. He started teaching me to do software development. Like, you have to run the service locally first. Oh, really? How? (So far nobody can run this specific service locally.) Or you should run it in a separate container. Oh, really? Do we have any documentation? Karen wrote some, but it's unsearchable, and Karen is in Japan, will be back Friday.

So I'm still negotiating with this guy. OMG how pissed of I'd have been just several years ago. No I don't fucking care, absolutely. The guy also admitted that he would not mind reviewing a 20-file pr with all the functionality included. Nice. Good to hear. I'll remember that. 

What else? Nothing. No, the service does not run locally, but it's a good challenge. I'll rather write an integration test that starts the service locally and pings it with different calls. 

Fuck, how come I have to listen to all these people all the time? I don't know. Maybe I should retire? Seriously.

Also, had this discussion:

chris 9:58 PM to be honest i think your approach tends to be too academic and not very practical. at the end of the day this isn't an academic exercise, we're trying to build software that meets a business need, so focus more in that direction. if the endpoint you built doesn't do exactly the right thing the first time, we can always change it later, especially for internal endpoints.

9:59 if you don't feel the code you wrote is even worth running before you submit a PR, don't submit a PR yet.

Juan

  10:03 PM

An interesting opinion. Thank you for this. I probably should have shared with you my experience in the industry, but anyway, when I hear about academic approach, I really feel good.

chris 10:04 PM

 

but the point is not to engage in academic exercises here.  save that for the university. (edited) 

Juan  10:05PM

:)

P.S. Update. Just unlinked from Chris Friend, May 22, 2024. Fuck them, these people. 

 

discussion at work

chris 9:58 PM to be honest i think your approach tends to be too academic and not very practical. at the end of the day this isn't an academic exercise, we're trying to build software that meets a business need, so focus more in that direction. if the endpoint you built doesn't do exactly the right thing the first time, we can always change it later, especially for internal endpoints.

9:59 if you don't feel the code you wrote is even worth running before you submit a PR, don't submit a PR yet.

Kirsha

  10:03 PM

An interesting opinion. Thank you for this. I probably should have shared with you my experience in the industry, but anyway, when I hear about academic approach, I really feel good.

chris 10:04 PM

 

but the point is not to engage in academic exercises here.  save that for the university. (edited) 

Kirsha  10:05PM

:)

P.S. Update. Just unlinked from Chris Friend, May 22, 2024. Fuck them, these people.