Issue #1: Start Me Up

🎶 If you start me up, I’ll never stop. 🎶

The Rolling Stones’ iconic line feels like the perfect anthem for the world of metaframeworks—a space that never stops evolving, innovating, and pushing the boundaries of what’s possible on the web. As we kick off 2025, the momentum from last year’s breakthroughs shows no signs of slowing down. From game-changing releases to bold new directions, the metaframework ecosystem is buzzing with energy. Let’s dive into the highlights, lowlights, and everything in between from the past year—and take a peek at what’s coming next.

The Good

Today, we have a bunch of interesting retrospective articles that provide a great overview of what various technologies achieved over the past year, as well as their plans and directions for new developments in 2025.

• The Astro team wrapped up the year with an awesome blog post that could even convince a skeptical old-timer to jump on the Astro bandwagon. Astro is arguably the most dynamic and innovative metaframework on the scene (though I’m pretty tired of updating my pet projects every month). With each release, the team expands its capabilities, and what was once associated with building static sites can now be used to create and deploy virtually anything that needs to exist on the web.
• The Svelte team delivered one of the best advent calendars I’ve seen in years, featuring daily news and updates about Svelte, SvelteKit, and the surrounding ecosystem. These updates were also nicely summarized in their first blog post of the new year. With the epic release of Svelte 5 in 2024, they have a lot to be proud of.
• Nuxt 3.15 saw the light at the end of the year, bringing developers some great DX (developer experience) enhancements. Nuxt has always been a framework with a strong focus on DX, and this release is no exception. This time, the team introduced support for the latest Vite version, significant HMR improvements, and a host of other holiday gifts for fans of one of the oldest metaframeworks around.
• The Analog team took a big step forward in metaframework innovation by adding support for Analog Server Components in their freshly published release 1.11.0. Looking at all the additional DX improvements they’ve made, I can’t help but admire the team and what they’re doing with Angular, pushing it into the world of trendy metaframeworks like Next.js, Nuxt, and Astro—and even taking a step forward with some original solutions.

The Bad

Amid all the good and perfect news from the past year, the fate of the Enhance framework went almost unnoticed. However, Enhance is (or was?) one of the tools that pushed the minimalist, progressive enhancement approach to web development forward, building its innovations on native web platform primitives. Now, things are pretty quiet in the Enhance universe, as often happens after big acquisitions. Still, I truly hope this isn’t the end and that we’ll see further developments, either on the Enhance side or with Architect. The world of metaframeworks desperately needs projects like Enhance.

The Noteworthy

After tinkering with Bolt and Lovable—tools for building full-stack applications in a single browser window—I came across an insightful research piece by Anthony Sidashin about his personal experience with both tools. It was interesting to learn that both projects have open-source implementations, allowing you to dive deeper or even use them on your own. I believe such technology will win over more and more developers in 2025, making simpler tools like v0 and IDX obsolete. The promise of metaframeworks, enabling the creation of full-blown web apps with databases in hours rather than days, is simply too tempting. In my experience, Lovable offers a slicker experience geared toward less technical users (which is nice in its own way), while Bolt gives users more control over the bare-metal aspects, which will likely win over seasoned web developers. Especially since Bolt isn’t limited exclusively to shadcn-ui and React.

Speaking of React, Tanner Linsley shared his own perspective on the future of metaframeworks, where (unsurprisingly) his brainchild, TanStack Start, takes the lead—at least in the React world. This is a very plausible outcome, given the trust and popularity TanStack tools have earned over the years. I particularly love the approach TanStack Router takes with its emphasis on maximum type safety, something other frameworks still struggle to match. This seems to be part of a broader trend of moving away from the [arguable] fragility and rigidity of file-based routing, and if so, TanStack Start is undoubtedly at the forefront of this movement.

As you can see, the year has started with a lot of exciting developments and even brighter hopes. And as usual, it’s hard to ignore metaframeworks if you’re even remotely involved in web development these days.

👋