I’m moving on to something new

[tipiirai]

Hey everyone. I want to share something I’ve been thinking about for a while now. I’m stepping away from Nue to work on a new product. This wasn’t an easy call, but I think it’s the right one. Here’s why.

The ecosystem chose React and Tailwind

I spent two years making the case for a simpler, standards-based alternative. Better architecture, faster builds, less complexity. None of it mattered. The frontend world picked React and Tailwind, and that’s not changing. I could keep fighting that fight for another two years, but I’d rather spend that energy on something with real traction.

AI changed the game

Development is moving fast towards agentic workflows. Developers aren’t choosing tools the way they used to. They’re building with AI, shipping with AI, and the old arguments about framework architecture matter less every month. Even Tailwind is in trouble, as developers stop visiting their site. A new frontend framework has almost no chance in this environment. In fact the whole software industry is in trouble: developers don’t read docs anymore, they prompt. They don’t browse landing pages, they ask an agent to pick the best tool. Discovery, adoption, learning, all of it runs through AI now. The traditional playbook of building an open source community, writing great docs, and growing word of mouth is losing ground fast. The tools that win are the ones that fit into agentic workflows, not the ones with a carefully crafted documentation or README.

I found a better way to do what I love

The thing I care most about is design engineering, particularly CSS. That hasn’t changed. What changed is that I found a way to express that love in a form that’s more useful and more aligned with where developers are heading. Still minimal, still rooted in web standards, but now built around AI and what developers actually need today. I’m quite happy I found the perfect niche to express myself, as the space for new developer tools is shrinking rapidly. I’m still confident that CSS is the way forward, not the weird idioms that come with the React stack. Agents need a direct description of what things look like, not six layers of abstraction. And honestly, developers always needed that too.

If you’ve been following my work because you care about CSS, clarity, performance, and doing more with less, I think you’ll like where this is going. I hope you’ll stick around and stay on the mailing list. I have an absolute banger coming and I can’t wait to share it with you.

With ❤️

— Tero

[Sharafudeen]

@tipiirai
Waiting almost 6 months for the next version of Nue.js 3!

To me, Nue.js is not just another web technology — it’s a true revolution. It empowers developers to build multi-tenant applications and websites with simplicity and efficiency. More importantly, it gives opportunities to people with limited resources to innovate, compete, and stand alongside the biggest tech giants.

Proud to be part of this journey and excited for what’s coming next! 🚀

[yanghuidong]

I hope you’ll stick around and stay on the mailing list.

@tipiirai You talking about the mailing list on nuejs.org, or somewhere else?

Just found out about Nue yesterday, totally randomly, via JS Rising Stars 2025. Yeah, AI, talk about "disruption" for real : ) But I think this has been a very meaningful endeavor of yours, and very thought-provoking, inspiring, even. AI has become very knowledgeable and capable indeed, but humans still have an edge on sense and taste. Looking forward to your sharing about the new project!

[richard-hajek]

That's a shame, I loved the direction you're going in! Is there another alternative with similar ideas as Nue is? Do you think Nue is mature enough to be used regardless of you choosing to focus on different projects? Is the code, in your opinion, elegant enough that if a problem would arise, someone with no experience in Nue (but with experience in JS and open source) could just fix it in Nue itself with reasonably little trouble?

[andreyluiz]

@tipiirai Really sad to learn that. But honestly, I think it's a good move. You saw the AI trend eating ground and decided to move forward, even with a repo with 8k stars. That's a remarkable decision that showcases your maturity and industry awareness. More than a great initiative, you built a community of people who admire you exactly because of this vision. Keep it going. We'll always be around.

[nonfungibletunji]

Any news on the banger, agentic tooling update.

[KazimirPodolski]

Just found out about this. Sad, I was waiting for Nue to mature to finally jump onto.

I could keep fighting that fight for another two years, but I’d rather spend that energy on something with real traction.

Ummm, why?.. There is no reason in the world forcing you to fight React/Tailwind/etc to the death. Niche projects are all over tha place, and while plenty of them are dead, many are thriving in their own field.

The tools that win are the ones that fit into agentic workflows, not the ones with a carefully crafted documentation or README.

Again, how comes there is nothing between "take over the world" and "discontinue the project"...

Anyway, @tipiirai, announcing discontinuation on the website and archiving Github repo would be the easiest courtesy to the community and unsuspecting users.

[tipiirai]

Ummm, why?

Why spend time on something with real traction?? I feel the answer is obvious: a project without users sounds useless and Nue wasn't "thriving". Where did you get the impression that taking over the world was the only option?

Not discontinuing yet, since the stuff that I'm building might actually end up being Nue 3.0. Not sure yet.

[mitorun-studio]

Personally, I really want NUE to be created as the simplest and most functional framework for creating basic static websites out of the box. Like version 2.0, but with everything needed for creating an average website: image and SVG compression, automatic image generation for OG, custom scripts for adding individual components (sliders, burger menus, lightboxes, modals, etc.). In other words, create the perfect tool for beginners who only need this functionality. Eliminate all the bugs so that it becomes a fully-fledged framework that will be maintained and gradually updated, without requiring major changes. Create a finished product. And then, using it as a base, you can develop a more complex framework for applications and AI.

[tipiirai]

100% agree. This is still on my mind.

[mitorun-studio]

It's been a year since I tried NUE 2.0, but I'm still impressed and can't wait to use it again. Back then, I truly enjoyed it; every morning I woke up wanting to create something in NUE. Unlike Astro and other frameworks, which either inspire indifference or even disgust. I'd highlight two development goals for NUE that can be used to attract users:

1. You'll enjoy working with it.
2. All the necessary features out of the box.
   Other advantages: extreme ease of use, incredibly simple setup, and web standards compliance.

I would already be using NUE 2.0 as my primary website building tool if it were to fix bugs and be confident that this product would be functional and supported for the next five years.

Ideally, the following should be added:

• Remove all text describing NUE's advantages and repetitive text from the documentation – move this to the main page or product page. The documentation provides a clear description of all the functionality and numerous examples.
• Add raster image compression. For example, based on Sharp. Ensure that <img> and ![]() image notations in Markdown remain unchanged, and all compression settings are in the configuration file.
• SVG optimization with SVGO. Assemble SVG into a stack sprite.
• More options for creating custom variables and links to social networks in the configuration file. For example, . Or, you can specify a link to a social network (or anything else) only once in the configuration file and insert it on the website as a variable: <a href="mailto:{ links.email }">{ links.email }</a>.
• Automatic generation of OpenGraph meta tags and images.
• Automatic sitemap.xml generation, including multilingual sites.
• Breadcrumbs.
• Create multilingual websites.
• Support for light and dark website themes.
• Many examples of working with collections: displaying cards with various filters and sorting.
• Automatically adds target="_blank" rel="nofollow noopener noreferrer" attributes to external links (required for MD).

If you want even more:

• Custom components with and without JS: burger, slider, modal, popup, spoiler, lightbox... Each component has its own separate JS code. And the styles are completely customizable in CSS, with examples.
• Custom CMS out of the box. For inspiration, check out Decap CMS and Keystatic CMS.

[mitorun-studio]

Give me a finished product, and I will use it and recommend it to others)))

[dumindu]

@mitorun-studio OG image generation can be a simple simple rust binary, go through pages in the project and generate according to front matter. This is quite dirty as I went thought different libs but very simple script; https://github.com/learning-rust/learning-rust.github.io/blob/main/image_generator/src/main.rs

I was a very happy customer of GoHugo but I love Nuejs v1 demos created with real world large datasets/ slider based forms. Nue can be like Hugo + Backbone.js full power of static site generation based on Markdown & yaml + frontend framework.