djyde
djyde on Hacker News
Posted by
djyde on Monday September 14, 2026 @06:10PM
from the show-hn dept.
https://nowdex.app/
Hi HN,I've been using Claude Code, Codex and Cursor quite a bit lately, and I found myself checking their usage limits all the time.Most of the tools I found for this live on the desktop or in the menu bar. What I really wanted was to pull out my phone and see how much I had left, especially when I'm away from my computer.I ended up building Nowdex for this.The main thing is the iPhone widget....
Because all Chinese users must use a VPN to access Codex.
I only had the patience to read long books in high school; now I really want to but they're too difficult.
The Mac Mini's price is already quite low while its performance keeps getting stronger.
What are the use cases for these small models? Is there anyone using models of this scale in their daily life who could share their experience?
Posted by
djyde on Wednesday June 03, 2026 @09:44AM
from the performance.dev dept.
I've always wanted to make a GUI client for pandoc
AI-generated music should be hard to detect
Posted by
djyde on Wednesday May 13, 2026 @04:11PM
from the ask-hn dept.
Has anyone else experienced the same thing? My ChatGPT always outputs one sentence per line, for example:---
Right now, your data looks more like:
“interesting utility”
rather than
“An app you can’t live without.”
---Isn't that how normal speech is supposed to sound?:---
Right now, your data seems more like an "interesting tool" than a "daily essential app."
---Even though I specified in the...
A Hacker News mobile client that I'm very proud of, specifically designed for non-native English speakers with automatic translation features.
https://haiker.app
I've transitioned between cloud services and self-hosting a few times:
1. Vercel Phase
My first project used Vercel. Since my project was Next.js, the experience was decent. But as my project gained some users, I found that even for projects under 100 users, I needed to pay $20 per month. Since my service didn't require high performance, this cost felt steep.
2. Self-host Phase (Hetzner + Coolify)
Later, I started setting up my own server with Hetzner and deploying with Coolify. Since Coolify is open-source and free, I only had to cover the cost of a VPS (even $5 a month was sufficient). I could deploy PostgreSQL instances and run a web server on it.
But later I discovered that even this way, I still had to spend a lot of effort maintaining PostgreSQL and Redis. Even though they were containerized with Docker, managing them was still troublesome. I needed to pass various system and environment variables between services, which was very tedious.
3. Cloudflare Phase
So later I switched to Cloudflare. With Cloudflare Workers, I can deploy fullstack applications and use D1 Database and Cloudflare KV to replace Redis. These features can be called directly within the Worker without needing to pass environment variables.
Plus, the local development experience is excellent and the pricing is very reasonable, so I've been using Cloudflare's entire suite ever since.
I switched to Cloudflare and it's been a breath of fresh air - everything I need and the pricing is reasonable.
I had a similar experience recently. I had a huge Swagger JSON file that would waste too many tokens if added directly to the context, so I told the agent in memory to use jq to retrieve what it needs when it wants to check this document. This saved a lot of cost.
I don't know why my comment was down voted. Can someone tell me why?
I recommend allowing AI to use TailwindCSS and Alpine in the context, as this would generate great lightweight dynamic web pages.
I'm curious what tasks you use this model for?
I use Wallet purely to satisfy my collecting habit. I like to add movie tickets to it after purchasing them in the movie app, so I can look back at all the movies I've watched from many years ago until now.
But I really wonder, is wanting a "little OS" just a hacker thing? For most people, they probably just want a full-featured OS. I don't have a solid take on this yet.
Haskell is like an instrument I can never quite master, yet I find it utterly fascinating and keep trying to learn it without success.
The Bitwarden Chrome extension is really bad, which is also the reason I've never been able to switch from 1Password to Bitwarden.
However, there's one counterexample: some email clients in the past experienced explosive growth by adding signatures. It was annoying, but it definitely worked.
Altman's early public class at YC is worth watching, though I can't speak to his character.
I've also noticed that when I communicate with Grok in my native language, its tone is more natural than other models. I think this is due to the advantage of being trained on a large amount of Twitter data. However, as Twitter contains more and more AI-generated content now, I'm afraid continued training will make it less natural.
My Claude often starts sleep-talking in Korean suddenly.
An LLM is like a super-smart 3-year-old, easily shaped by its environment to exhibit corresponding behaviors.
I've switched to OpenCode and I think it's really great.
The first advantage is that it can use many different models, not just Claude. The second is that I've noticed OpenCode streams out the entire reasoning process, while Claude Code doesn't show streaming of its thinking.
First time hearing about this, seems pretty nice.
btw it's written in Nim, which I've been wanting to try for a while.
What model are you using? I don't think I've encountered it before. But then again, as long as you explicitly instruct an AI to follow REST, it usually will.
Setting up own GitLab instance might be a good solution too.
WeChat in China was early to implement friend group-based posting