Hacker News new | ask | show | jobs
by chmln 2187 days ago
That document is not particularly enlightening either.

> we're removing the types from internal code and making it pure JS. this reduces complexity and helps us ship a faster product

Yeah, because dropping static types from 10K lines of code is definitely not going to be a maintenance nightmare in the long run.

But perhaps this is something beyond the simpleton comprehension of us novice programmers
2 comments
klodolph 2187 days ago
> Yeah, because dropping static types from 10K lines of code is definitely not going to be a maintenance nightmare in the long run.

My personal experience is that there’s a size threshold where static tying becomes more useful. Below this threshold, the problems solved by static types are still mostly tractable by humans and things like linters. The threshold is different for every program and set of developers.

10K lines is small enough IMO.

> But perhaps this is something beyond the simpleton comprehension of us novice programmers

That’s just flamebait, the comment is better without it.

link
tome 2186 days ago
It was your GP who introduced the word "novice".
link
klodolph 2186 days ago
It’s not the word, it’s the sentence.
link
inimino 2187 days ago
> That document is not particularly enlightening either.

Here are some key words that many kibitzers here could benefit greatly from internalizing:

> > The target audience is the 5 to 10 people who work on this particular internal system.

> Yeah, because dropping static types from 10K lines of code is definitely not going to be a maintenance nightmare in the long run.

Are you one of the people responsible for maintaining those 10KLOC?

> But perhaps this is something beyond the simpleton comprehension of us novice programmers

I believe Ryan Dahl would admit to having made more mistakes in large system design than most of us will ever have a chance to. Who knows, he may have picked up a thing or two along the way that most of us haven't.

This could also just be an internal product decision that really doesn't affect anyone outside the core project team, while some of us here are projecting way too much onto it.

link