Hacker Newsnew | past | comments | ask | show | jobs | submit | dcreater's commentslogin

People have been saying this before chatgpt and ever since. And they're right.

Its charlatans like sama that muddy the waters by promising the sky to get money for their empire building.

LLMs can make and are great great products. But its sneaky salesmen that are the ones saying scaling is the path to AGI. The reality is that they're just aiming for economies of scale to make their business viable


Fixed title: Python is not a great language for data science if pandas/polars/ibis did not exist

Please read the article. It literally shows pandas code as an example.

That's good to hear!

If the health monitoring is reliable, im sold. I want to move on from apples clutches despite the pebble hardware not looking particularly attractive to me


It worked and it was good enough

At 1Mbit. It was good enough but it absolutely sucks today. Meanwhile AirDrop is hundreds of megabit to a gigabit.

Trying to send a video file over Bluetooth would be miserable.


What are you actually arguing here?

You're saying as if it's a bad thing. We need new blood. Thank god IBM Failed and we got microsoft /hp/apple. Now it's time for the next wave to make the current Dinos go extinct


Its specifically meant for control systems no?

hardware engineering is a very broad field and the title is misleading


This is my project/blog post and yes, I've gotten that feedback a couple of times. I mostly work in control systems and probably reflexively think narrowly about that context. But it's definitely tailored to controls - I wasn't intentionally trying to oversell it or anything.

It’s the title of the blog post and I didn’t want to change it. But yes, it seems to focus on the specific subset of hardware engineering that’s control systems.


It's C codegen using casadi under the hood. Most embedded systems can compile some form of C.


Yes!


Looks vibe coded?


The CLAUDE.md file is right there, so they are probably using agentic coding.

But why does it matter? Does the app not work? I don't have a Mac, can't check.


Counter question: How do you know it works?

A file manager better be rock solid, I don't want a bug to delete any files or do other shenanigans.


That is a valid question.

But that would apply to any app that deals with files like this one does.

This one is open source and we can run some code analysis on it, compile locally, etc. I am not well versed in security checks but I guess you get the idea.


Especially the 'design' which couldn't be less Mac-native if it tried



Of course it is.

As usual, it is low quality and has zero tests.


This situation reminds me of the Visual Basic apps in the nineties. It was possible to write good quality software with VB and some people did. Nonetheless the majority was dross and the fact that an app was written in VB was a strong indicator of poor quality.


agents should agentically create high quality unit tests


Agents are really good at writing unit tests, but only if you clarify exactly WHAT should be tested. Otherwise they generate slop which passes, but don't catch any bugs nor regression.


Ugh. This term can't die fast enough.


What term would you suggest instead? Agent-assisted slop?


Not bad; anything that actually mentions "AI" would be better than "vibe." That doesn't tell you anything.


Why do you say that?


Because it is true. Claude did almost all of it.


It appears very "vibe coded", the application makes use of the stereotypical AI gradient in the top bar of the application.

The README.md uses a ton of Emojis in the feature Setting.

100% AI.


Don't know why this comment was flagged, this is exactly what is going on here. AI is famous for these purple gradients and the website reeks "vibe-coded", this is 100% how Claude makes websites, including the already mentioned gradients, emojis, style, etc.

And of course every commit is Co-Authored by Claude Code with excessive commit descriptions also created by Claude. Is this really something we want to see on Hacker News? I wouldn't trust such an application.

Nothing against AI coding but letting AI take the wheel 100% of the time and not even mention it (like he coded it himself) is very dishonest.


Vibe coded thing number 367 in this category


Following up the previous post, made a v2 collating popular best practices/guides found on HN, reddit etc. Improving the process to address some of the shortcomings last time (primarily separating out general software engineering best practices - still critical and foundational, but separating them out allows more clarity on agentic coding specific best practices)

https://rosmur.github.io/claudecode-best-practices

TOC: 1. Executive Summary

2. Sources

3. General Software Engineering Best Practices

3.1. Test-Driven Development (TDD)

3.2. Continuous Quality Gates

3.3. Code Review - Including AI’s Own Work

3.4. Incremental Commits with Clear Messages

3.5. Monorepo Architecture

4. Core Recommendations by Category

4.1. Context Management (Most Critical)

4.2. Planning \& Architecture

4.3. Tool Usage \& Automation

4.4. Workflow Optimization

4.5. Production Code Quality

4.6. Advanced Patterns

4.7. Testing Authenticated Routes

4.8. Architecture \& Design

5. Contradictions \& Trade-offs

5.1. Skills vs Context Bloat

5.2. Custom Subagents vs Clone Pattern

5.3. Auto-Formatting Hooks

5.4. Planning Mode vs Manual Plans

5.5. Documentation Volume

6. Appendices


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: