nzach

7 ms·

61.
by nzach 1y ago
I've got the same error on my side. At first I thought it was some weirdness with Firefox, but opening on Chrome gives the same result. I don't know what causes this bug specifically, but encountered similar behavior when I asked
62.
by nzach 1y ago
It is mostly because it creates code that is way more complex than it needs to. One, admittedly silly, example is Claude trying to put a light/dark theme switcher when you are trying to refactor the style of your small webapp. I'm
63.
by nzach 1y ago
>Provide detailed specifications for the work to be done I've been playing around with vibe coding for a few months and my experience doesn't really match this advice. I used to think this was the correct way and based on that
64.
Ask HN: Are LLMs changing how you use and contribute to open source?
3 points by nzach 1y ago|2 comments
65.
by nzach 1y ago
Care to elaborate on how you use the status.md file? What exactly you put in there, and what value does it bring?
66.
by nzach 1y ago
https://docs.anthropic.com/en/docs/build-with-claude/prompt-...
67.
by nzach 1y ago
In my experience it works better if you create one plan at a time. Create a prompt, make claude implement it and then you make sure it is working as expected. Only then you ask for something new. I've created an agent to help me create
68.
by nzach 1y ago
I agree with the sentiment, but I don't think the example given (creating SQL queries) is a good representation of this problem. That's because if you know a little bit of SQL and know how to validate the answer LLMs give you this
69.
by nzach 1y ago
One interesting thing I noticed in these "fixing bugs" demos is that people don't seem to resolve the bugs "traditionally" before showing off the capabilities of this new model. I would like to see a demo where they
70.
by nzach 1y ago
> For some reason people are willing to produce design documentation.... I'm assuming you wrote that just for dramatic effect but let me explain why I think this behavior is completely rational. If you implement "feature X&quot
71.
by nzach 1y ago
I'm starting to think about moving to a new city and was considering to build a tool like this. And now I'm excited to see this already exists! For my use-case the interface you created isn't the best. Now that I'm searc
72.
by nzach 1y ago
I'm working on https://leetprompt.io It is a platform for prompt challenges, in leetcode style. But you act as a director for the AI writing the code. You ask what you need and the AI implements your request. The idea is to
73.
by nzach 1y ago
> won’t let the program even compile if there is an unused variable That is a really big advantage in the AI era. LLMs are pretty bad at identifying what is and what isn't relevant in the context. For developers this decision is pre
74.
by nzach 1y ago
> It can show you good idiomatic code in small snippets. That's not really true for things that are changing a lot. I got a terrible experience last time I've tried to use Zig, for example. The code it generated was an amalgama
75.
by nzach 1y ago
> How is this not excellent advice? I do understand that in order to create something popular you need to create something good but you also need to properly communicate what you do. And proper communication is as hard as creating someth
76.
by nzach 1y ago
> Is there a website, documentation, any kind of presentation of your product? I do have a fully functional MVP available on the internet ( https://leetprompt.io ) > The other side is, people are also so feed up with seeing
77.
by nzach 1y ago
I just want to share my recent personal experience. Recently I've finally decided to try creating something new that people would find useful hoping that some day I would be able to turn a profit from that. So I vibe coded a pretty bar
78.
by nzach 1y ago
Aider is in large part written by AIs, they even have some stats here: https://aider.chat/HISTORY.html Besides that, I think your best bet is to find someone on youtube creating something "live" using a LLM. Somet
79.
by nzach 1y ago
> take several minutes to do something The quality of the generated code is inversely proportional to the time it take to generate it. If you let Claude Code work alone for more than 300 seconds you will receive garbage code. Take that a
80.
Show HN: LeetPrompt – Leet Code for Prompts(leetprompt.io)
1 points by nzach 1y ago|0 comments
81.
by nzach 1y ago
Is there a proper way to use this project without commiting files to git ? I just want to try it out in a project I'm working on, but don't want to put it in the history. What I did is to add the backlog folder into the .gitignore
82.
by nzach 1y ago
I agree with the sentiment of this post. I my personal experience the usefulness of a LLM positively correlated with your ability to constrain the problem it should solve. Prompts like 'Update this regex to match this new pattern'
83.
by nzach 1y ago
I'm experimenting with a somewhat similar workflow. I have a Boox tablet I use to take handwritten notes on any relevant topics during the day. And with that I create the following workflow: 1. transform handwritten notes into a markdo
84.
by nzach 1y ago
> What explains the slowdown in IANA timezone database updates? My guess is that with the increasing dependency on digital systems for our lives the edge-cases where these rules aren't properly updated cause increased amounts of pai
85.
by nzach 1y ago
> "Are you presuming most people are stupid?" I really like this approach, but I tend to use a slightly modified version that is along the lines of "explain me why an intelligent person would do [dumb thing]". I think
86.
by nzach 1y ago
> Heck, even MCP itself isn’t new—the spec was released by Anthropic in November, but it suddenly blew up in February, 3 months later. Maybe it was because OpenAI announced they would start to support MCP in their tools ? [0] Perhaps I&#
87.
by nzach 1y ago
This seems to be a better alternative: https://github.com/mark3labs/mcp-go
88.
by nzach 1y ago
And if you "reverse" this idea you can make a "holographic(?) display"[0]. [0] - https://www.youtube.com/watch?v=ric-95ig5oE
89.
by nzach 1y ago
Isn't 'doctors political beliefs' just another axis(proxy) people can use to make a guess about the competency level of said doctor? If you 1 - think X and also 2 - think that everyone who doesn't hold the same belief is
90.
by nzach 1y ago
> why we not just use API Did you meant to write "a HTTP API"? I asked myself this question before playing with it a bit. And now I have a slightly better understanding, I think the main reason was created as a way to give acce

More ›