HNHacker News
TopNewBestAskShowJobs

gulugawa

55 karma · joined January 15, 2026

gulu@createthirdplaces.org
submissionscomments
gulugawa··on You don't need React: creating a minimal UI library in Vanilla JavaScript
The blog post page, which appears to use the author's library only loaded 58.6 kB of JavaScript, which is quite low.
gulugawa··on You don't need React: creating a minimal UI library in Vanilla JavaScript
I looked at the author's website, and a typical page downloaded around 60kb or less of JavaScript.
gulugawa··on You don't need React: creating a minimal UI library in Vanilla JavaScript
I think the high quality of the author's website is more relevant than theoretical concerns about performance.

Loading performance on the s website is quite good, and it appears to be using the JS library described in the blog post. I also saw that pages were loading around 60kb or JavaScript or less.

Also, from looking at the author's GitHub profile, it appears that they do not make significant use of LLMs and actually took the effort to write the code by themselves.

gulugawa··on You don't need React: creating a minimal UI library in Vanilla JavaScript
I read the blog post, and your framework looks great.

I think React is overused, and I'm happy to see people support alternatives.

gulugawa··on What should go in your Claude.md?
Claude.md files should include a dedicated language for talking to computers.

There is too much ambiguity in English that makes it a poor choice for giving a computer instructions.

gulugawa··on Zitron: "Everyone Has Been Sold a Lie" on AI [video]
He's right.

LLM AI tech is a scam.

gulugawa··on Hacker uses DeepSeek AI to autonomously attack vulnerable servers
This could be easily solved if DeepSeek had clear boundaries on what it could do.
gulugawa··on Rogue AI hacking incidents amplify debate over open-source tech
This seems more like a problem with poorly developed software that has no clear boundaries. With basic security measures, developers will have to out of their way and intentionally create automated hacking tools.

I develop open source software, and I am very specific on what it is able to do. The software doesn't connect to any external dependencies expect for specific services used for authentication and data persistence.

gulugawa··on Can I vibe code it?
All the site needs is this one line of html. <h1>No</h1>
gulugawa··on "Google and Reddit do not own the Internet," web scraper says after court win
My websites are not as big as Google or Reddit, and I take steps to block bot scraping.

I'm not interested in paying bandwidth costs for bots that don't care about what I posted.

gulugawa··on Mark Zuckerberg Blasts Centralization of A.I. Power
Who is this new Mark Zuckerberg?
gulugawa··on EYG: A Programming Language for Humans
Good luck. I look forward to hearing about future progress.

I strongly agree with the approach of making programming more accessible with a simple, deterministic feature set.

gulugawa··on Ask HN: What apps are you building?
I built a mini frontend framework with components, state management and backend data fetching.

It is designed to have a minimal learning curve, and has faster performance than React due to the use of web components.

gulugawa··on Removing React.js from the codebase and adapting Htmx for UI interactivity (2023)
One key part of good frontend engineering is picking tools to effectively solve current challenges your product is facing. Thinking about theoretical problems that may not apply to you is a waste of time.

I don't currently use use HTMX. However, that doesn't invalidate the use of htmx by others when it is a better fit for their experience and use cases.

gulugawa··on Removing React.js from the codebase and adapting Htmx for UI interactivity (2023)
The overwhelming majority of websites are small projects that don't need complex tools to provide interactivity.For those websites, complexity is more likely to come from planning for complexity that will never exist.

Using a tool such as htmx is an effective way for them to solve current challenges while minimizing complexity.

gulugawa··on Removing React.js from the codebase and adapting Htmx for UI interactivity (2023)
I designed my own mini-framework Places.js that you might be interested in looking at. https://codeberg.org/createthirdplaces/places-js

Places.js supports component based rendering, and state management integrated with backend data fetching. It also has faster rendering performance than React due to the use of vanilla JS web components.

Places.js is also designed to dynamically render parts of a page while making sure the rest of it is statically rendered. HTMX can be used on those parts, and be used to enhance HTML generated by components through the htmx.process() method. https://htmx.org/api/#process

gulugawa··on Removing React.js from the codebase and adapting Htmx for UI interactivity (2023)
An islands architecture can be implemented with web components when hydration is needed.

Writing vanilla JS to manage state and backend data loading can be done with a script that is less than 2kB zipped. The script can then be reused.

gulugawa··on Htmx 4.0, the first JavaScript library to release exclusively on the Game Boy
https://htmx.org/essays/paris-2024-olympics-htmx-network-aut...
gulugawa··on Htmx 4.0, the first JavaScript library to release exclusively on the Game Boy
While I prefer to stay with an approach of using web components for my current work, HTMX seems like a good option to consider for future work.

I think the design of HTMX acts as a complexity limiter by encouraging less JS. One can make the argument that manually writing JS instead of HTMX would be easier and more maintainable. However, I've found that JS code often trends towards becoming increasingly complex over time.

Also, I'm glad to see that htmx can be installed without npm. I think the ability to install npm packages can incentivize too many dependencies.

gulugawa··on The drivers behind software delivery inefficiency
Great paper. QA's role in software is underrated.
gulugawa··on When will language models be good enough?
Never
gulugawa··on Show HN: A Pipeline for Making 10-minute AI Movies with Claude Code and Seedance
Its going to cost a lot more than $200 once LLM subsidies run out.
gulugawa··on LLM-Integrated Multivariable Calculus Course
I asked the bot questions about competitive Pokemon and Game of Thrones which it answered.
gulugawa··on Postgres rewritten in Rust, now passing 100% of the Postgres regression tests
I don't code with LLMs, and think you might be right.

However, Postgres is a tool with clearly defined functionality and doesn't have ambiguous requirements that are seen in user facing software. As a result, it is entirely plausible that the author created a working Postgres replacement for certain use cases.

I personally want to see more evidence about the quality of the tool after it is in a finished state.

gulugawa··on Postgres rewritten in Rust, now passing 100% of the Postgres regression tests
This the type of AI generated tool I'm interested in seeing more of. I'm a fan of how the quality is verified using clearly defined quantitative measurements for performance and correctness.

I'm still skeptical about LLMs and don't use them, although I can be convinced by more demonstrated examples of success.

gulugawa··on What is agentic AI today, and what do we want it to be?
Exactly, except that it wouldn't query models running on the public Internet.
gulugawa··on Virginia bans sale of geolocation data
As someone who frequently travels to Virginia, this is great news.
gulugawa··on GLM-5.2: The Open-Source Chinese Model Challenging Claude at One-Fifth the Cost
Interesting. Since GLM-5.2 can run locally, I'm wondering how challenging it would be to fine-tune multiple GLM-5.2 instances and run a collection of self-hosted agents.
gulugawa··on What is agentic AI today, and what do we want it to be?
I think agentic AI would be improved if it was a collection of smaller models that are working together. This has the potential to make costs significantly cheaper and improve accuracy.
gulugawa··on The Case Against Public Engineering Levels
I'm inclined to agree with the post, although it feels strange. In my experience at tech companies, not having public engineering levels improves collaboration and improves product quality.
← PreviousPage 2 of 4Next →