Tell HN: I kinda want to go back to Java

This logic makes no sense to me. Names aren't an inherent property of languages; the names you didn't like come from either someone else's "framework" or your own design decisions. Lots of Java code has short names (including single-letter iteration variables), and lots of Python code has long names — for example, the standard library `email.errors` has a class called `InvalidMultipartContentTransferEncodingDefect`. The number of things to name, meanwhile, is purely a consequence of the problem domain, how you model it, and how you factor the code.

"Give me a name for this thing" is the last micro-managing coding task I would assign to an LLM, because the entire point of the name is to express my own intent behind the code.

Besides which, you haven't identified something you actively dislike about Python, just that a problem you supposedly used to experience in Java is supposedly ameliorated now.

> The Context Dependency Injection stuff was so great, allowed creating things on demand and in different scopes of demands so easily.

OP is currently working in Python, so that seems to me like a wild justification to go back to Java.

Elsewhere ITT someone is praising Java for having the ability to keep large codebases organized and comprehensible. But my experience of Java was that the overall boilerplate, focus on classes, manifest typing, functions not being first-class objects, etc. make the codebase large. Then it gets ballooned even further with systems designed to introduce dynamism into a fundamentally static language.

As I've pointed out in other threads before, "dependency injection" in a dynamic language is approximately a fancy term for "passing an argument to a function". Oh, but it already looks for a global and you don't want to rewrite it for some reason (even though your compilation step is trivial)? That can be addressed with a pretty small decorator. It seems like a lot of why these frameworks exist in Java is so you can reuse implementations of "design patterns" that fix problems Python doesn't have in the first place. By the way (since I just checked it out of curiousity), the code for the Spring framework is comparable in length to the entire Python standard library, including tests.

I program primarily in Typescript and Golang these days. And yes, most corporate js people tend to not know much about things like dependency injection until they end up with the kind of problems that led to dependency injection in the first place. But Javascript/Typescript caters to everyone's taste and level. If you want to do enterprisy stuff and write properly constrained software - you get to do it.

Working in a language that tries to enforce "type" discipline and certain conventions really gets you down to solving problems the right way - instead of just hacking together a workaround. So most of the time, people who write Java tend to be more thoughtful about their approach to various things in a way that the typical Javascript developer probably isnt.

> Python helped us invent LLMs - it's quite possible if it didn't exist we would never had invented them.

Sorry, but this is nonsense.

I wish I could write C for $$$. My brain is too fried after work+kid for C. Yesterday I spent a couple hours to write a section of my meter to read literal strings without escaping chars. Should only take some 30m at most.