trevordev
trevordev on Hacker News
Posted by
trevordev on Tuesday September 22, 2026 @11:36AM
from the show-hn dept.
https://github.com/TrevorDev/niftyagent
I built this as I wanted to see how little setup is needed for a useful coding agent which I could use as a baseline for experimenting with different context management or task completion strategies.The whole thing is under 500 nodejs lines and has no framework or libraries as dependencies. Its basically a loop with a small set of tools to read/edit files.The interesting part to me is being...
I did 4 months ago, I loved it at first but when I tried to use it in a more complex use case during hackathon I ended up hitting some weird compile bugs with it and decided to go back to webpack. Maybe ill try it again once it matures a bit more.
Front end: Typescript, Vue, Bootstrap
Backend: Node, Typescript, Express, TypeORM, Postgres, SocketIO, PugJS, BabylonJS
Webpack for bundling
Digital Ocean or Azure for Ubuntu vm hosting
I feel having the same language for front/back end really reduces mental overhead.
Yep, its seems like theres a lot of directions the web can go in and I like the more experimental concepts shown in your videos. Looking forward to see how lensVR will turn out.
Hey Billy, I think I chatted with you over Skype a little while ago :). From what I've seen your team's teams effort seem the most inline with what I'm wanting but there are a couple things I haven't yet seen from lensVR:
- Desktop support (I have a Vive and I think you mentioned its only available for gearVR right now)
- Multitasking, all the images on your twitter seem to only have one site at once
- Fallback to 2D site (If Im in a webVR version of reddit and I click an unknown url I need to be able to multitask on
that site even if it doesn't have 3D support)
Since I have yet to find a vendor heading towards support for these things, currently Im using a fork of firefox where I just hacked out a bunch of security settings. Right now it lets me load iframes and pull content from them into webVR but it was just a quick way to get it working as I'd prefer if each separate tab could draw to its own layer and then be composed with other tabs (Im hoping this is part of WebXR standard but it sounds like it wont be there for a long time).
If you have anything else to share with me or if you have any additional questions for me, Id be happy to answer. I would LOVE to get a browser I could target so I can stop worrying about building my own browser and focus on my web apps.
Thanks!
Dang, sorry to hear that. I've seen those projects in the past but the problem I have with those is that they seem to only support web 2D multitasking using browser APIs. Here's hoping you reconsider in the future. Thanks
This project looks awesome. Do you know if you have any plans to push some wild/experimental webVR features into this? I am desperately looking for a browser that will support:
- webVR multitasking (>1 tab rendering to display a time)
- tab content sharing
- obtaining textures from html elements to project into VR
so I can watch twitch in WebVR while browsing Reddit in VR. So far I have a sketchy fork of Firefox that somewhat works but I'd love for something more stable to recommend to people. What I want is described in more detail here: https://docs.google.com/document/d/1JzxKqir57c4RnpHhpH8nED8C...
Let me know if you have any thoughts. Thanks.
I have a vive and what I don't understand is why I need to use chromium or nightly firefox to view this content when it is out on official mobile builds. Does anyone know why this hasn't made it into official desktop builds yet?
A mix of site's public api's, rss feeds and webcrawling
Anyone know why the CUDA toolkit is 1.2GB? It seems extremely large to get started with. In comparison Vulkan which is only 130mb.
Possibly the physics of the world is based on change in time between rendering. If the rendering time is slightly longer/shorter the collision calculations may cause this non-deterministic action.
I just tried to order one but got a 404 error.
Typescript has became my favorite language. The option of having types when you need them is great and the addition of async/await make working with promises a lot nicer. Ever since I've been able to use experimental features and get es6 support with typescript I've never looked back. The only thing lacking now is the tooling. The atom-typescript plugin is the best I've found but I would love to see better integration with other editors. Can't wait to see what comes to typescript next!
I really enjoy typescript and I really enjoy the cutting edge features. Atom editor is the only editor that has the typescript plugin that works with the latest features of typescript. I hate the performance of atom but it is worth the tooling.
I am not a fan of DoorDash's business techniques. DoorDash sent a flyer to my house false advertising a coupon code. I spent half an hour creating an order only to realize that the coupon they sent me was different from what the flyer said. When I called to complain they acknowledged their error but said they wouldn't do anything and that I should just place an order anyways.
I don't want this company anywhere near the last mile to me. DoorDash sent a flyer to my house false advertising a coupon code. I spent half an hour creating an order only to realize that the coupon they sent me was different from what the flyer said. When I called to complain they acknowledged their error but said they wouldn't do anything and that I should just place an order anyways.
2 months ago trying to use amazon fresh told me I needed to be a prime member to use it. I signed up for a year of prime for the sole purpose of using amazon fresh and I was very happy. Now 2 months in they are forcing me to pay another 200$. This may be my fault for not reading my prime subscription terms but I feel tricked into buying prime. I will be trying out instacart in the future.
Looks like they used string theory.
agreed but I prefer to use the same language for front and back end to reduce the context switch cost when working on both ends.
1) I don't think any of those tools are dying, people are just sharing that other tools will work better in certain situations and each tool has it's place. I'm a fan of nodejs but I have found the dynamic typing and poor performance of JavaScript can hinder my recommendation when building a large, performance critical app.
2) The languages/frameworks I would recommend depends on what you want to build. The large amount of JavaScript libraries and its ability to work in the browser make me highly favour javascript and I have found that typescript provides a nice layer on top to provide sane typing and code auto completion for many of my needs. However, Typescript tools and the language still feel underdeveloped. I would use node to handle web requests and offload "heavier" work to c/rust/go if needed
3) I used to use c/c++ for a lot of my personal project but I have found tasks like managing memory can add unneeded complexity to the majority of my tasks. C still has it's place when certain requirements must be met.
High caliber post. I'm curious about the manufacturing process and rate of firing error. There were a lot of moving parts and it would seem difficult to ensure that the gun won't get jammed or have other issues.
It looks like i am not being tracked but i am curious to how many people actually are.
I setup lets chat at a small start up I was at about 2 years and it worked very well then and im happy that it looks even better now. Will definitely consider it in the future.
Here is a android hacker news app I made to quickly get updates
Let me know your feedback
Features:
- Flashes HN headlines word by word to using Rapid Serial Visual Presentation.
- Tap a headline to view its link
- After a headline is read it won't be displayed again so only new content will be shown
Posted by
trevordev on Monday January 05, 2015 @08:22PM
from the show-hn dept.