Issue #21: We Are The World
I always like when something nostalgic pops up in the informational field of web development. It allows one to reflect on the path our industry has passed and ideate (or fantasize?) on the bright future we all may have. So this issue is largely about the retrospective research of the problems/solutions the JavaScript [meta]frameworks community came through/to during the last years, along with some news from different parts of the ecosystem.
The Good
I was really excited to read this deep piece about the React ecosystem from Mark “acemarke” Erikson — the person who definitely knows a thing or two about this topic. React, being a part of the most popular metaframeworks around, including Next, Remix, TanStack Start, and (optionally) Astro, and serving as an inspiration (in either a positive or negative way) for all the others, has become something ubiquitous on the web to the extent most building tools support it (or JSX at least) by default these days. There is everything in this article: good and bad parts, love and hate (some scandals overview included!), rants and ideas — but all in all, it gives a good perspective on where we are and where we are (and should be) heading.
This hopefully answers a lot of questions about how and why React has developed the way it has, what the influences are on React’s development process, the main goals of the React team, and where React usage patterns stand today. I also hope I’ve been able to dispel some of the confusion and FUD around the React team’s motivations and reasons for pushing in specific directions. It’s fine if people disagree with the technical directions, or decide that they don’t need React Server Components or to migrate to a larger framework. But the React team’s intentions are valid and genuine here. It’s really hard to maintain widely-used libraries and satisfy the variety of needs and usage patterns in the community. I do think the React team has done a good job overall.
The Bad
Last week, I stumbled upon the underestimated research (from Felipe Stanzani) on the positive and negative sides of Next.js and the options one has if they want to leave this popular place for a well-deserved rest. Interestingly, I have more or less the same thoughts on alternatives to Next.js myself, so it was a good experience for me to dive into, especially in terms of its depiction of the great work that happens behind the scenes with TanStack Start.
It’s important to emphasize that despite being developed by the same team that builds the entire suite of TanStack libraries, one of TanStack Start’s principles is that you are not obliged to use any of them. It’s an independent framework.
Notably, there was a good detailed comparison of Next.js and TanStack from the LogRocket guys too.
If you’re looking to move away from boilerplate-heavy workflows and want type-safe routing, smarter data fetching, and the flexibility to structure your app your way, TanStack Start might be exactly what you need.
Quite generic (and with a smell of light LLM painting for sure), as usual for the LogRocket’s team of technical writers, but at the same time insightful in terms of its side-by-side analysis, if you’re considering something like a migration, or choosing from these two options specifically.
Another notable thing that is always interesting for the JS community (even though it solidifies its flighty reputation) is the appearance of a new [meta]framework, which in general is good, even though it blurs the focus of curious developers each time. This one (literally) is based on our good friend React and brings the accent on the capabilities for cross-platform development, which is definitely a good (though not very new or original) and demanded niche in the modern world of short-span AI-driven vibe-software-craftsmanship. The One provides a good DX for developing with React and React Native (side-by-side), with some inspirations from existing solutions (like Remix, for instance) potentially simplifying migration. There are definitely some complexity tradeoffs (common for such flexible approaches, to be fair), but all in all, you’re doing something complex if you reach out to the One anyway.
The Noteworthy
Existing mainstream metaframework vendors also didn’t stand still in recent weeks and came up with a bunch of interesting milestones on their ways to perfection.
The Astro team published their beautiful round-number minor which introduces an interesting experiment like real-time-fetched live content collections, in addition to solidifying a lot of existing experimental features. The guys are preparing something interesting for us this week too, so I personally am all ears and all eyes. Meanwhile, if you’re into Astro too, check out the deep writeup on Astro integrations from Louis Escher — definitely something any advanced project needs to be aware of.
Analog got a new minor release, too, bringing improvements and fixes to component templates and aligning the tool better with the recent Angular 20 update.
The new technical update to SvelteKit brings support for Vite 7 and Rolldown betas — and that’s really cool that the guys let everyone play with the new tools that early (especially taking into account Rolldown being directly inspired and API-guided by Svelte’s homie Rollup).
Better metaframeworkization (sorry!) comes to Hono with its new minor providing decent enhancements for routing and lots of new enablements for different kinds of tooling integrations, bringing Hono even more of the flexibility it’s already quite famous for.
All in all, everyone is cooking something good this summer, and I bet September will bring a lavish amount of cool new releases, both in the React world and in the JS metaframeworks ecosystem in general. I’m also preparing something new and fancy for this newsletter, not related to any underlying tools but rather to messaging and additional goodies which will hopefully bring some value to the readers and the wider audience. So if you like this project, I would really appreciate you spreading the word.
đź‘‹