I mean, he's 33, that's well old enough to understand getting in over your head and understanding the value in humility, even if only deployed sparingly.
There's a reason back in the early Sony and iPhone days he earned the name "egohot", and seeing him come out of the woodwork when he launched comma just... validated all that all over again.
Highly unlikely. In most big tech companies you are strongly encouraged to begin pushing code immediately. You definitely don't need to know the whole stack from top to bottom to remove a popup which was what he was apparently tasked with doing (and didn't manage).
Quite the opposite. I've fixed bugs on my second day on a job when it was a simple, testable monolith - the way it SHOULD be for most companies still.
In a distributed system, a small change can have a giant blast radius. You don't know if the service you are touching was part of resilience testing. There are hundreds, it's easy to miss. What if you make this microservice throw all the time? What happens? Does it bring down anything else?
I don't understand why people still think that microservices make things easier. Distributed systems have always been insanely complex to reason about, and that never changed because Node or Docker, or something.
He even complained that it was impossible to get anything working locally because everything was so interconnected and complicated, so there is your answer.
Those are all good points that I hadn't considered.
The idea behind microservices is separation of concerns with horizontal scalability. The microservices should be independent enough that it's possible to test the integration in stages without bringing the whole system down.
My team relies on services offered by cloud providers, we are just composing them together, which means that we don't have to create userpools for example.
Hey, the point of an internship is to trial the job, during the trial he decided there wasn't much he could contribute and so he decided to bow out. That seems pretty reasonable to me.
17 comments
[ 3.2 ms ] story [ 51.1 ms ] threadfailing to remove a login popup- okay. internship- okay. He's young I guess.
https://en.wikipedia.org/wiki/George_Hotz
Guess to be productive there you have to stay at least half a year before anyone allows you to do minor changes on the core...
In a distributed system, a small change can have a giant blast radius. You don't know if the service you are touching was part of resilience testing. There are hundreds, it's easy to miss. What if you make this microservice throw all the time? What happens? Does it bring down anything else?
I don't understand why people still think that microservices make things easier. Distributed systems have always been insanely complex to reason about, and that never changed because Node or Docker, or something.
He even complained that it was impossible to get anything working locally because everything was so interconnected and complicated, so there is your answer.
The idea behind microservices is separation of concerns with horizontal scalability. The microservices should be independent enough that it's possible to test the integration in stages without bringing the whole system down.
My team relies on services offered by cloud providers, we are just composing them together, which means that we don't have to create userpools for example.
"Area man does thing"
> This is the attitude that builds incredible things. Let all the people who don’t desire greatness leave.
He must just not have desired greatness enough :(