Thoughts on Astro framework?

What would you say it’s good for? Have you used it before? Will employers be looking for Astro developers anytime in the future?

It’s absolutely fantastic for creating static page sites but with the convenience of a component framework. You can implement React or Svelte or whatever components for specific reactivity, which is great.

I’m a big fan. I wouldn’t use it for primarily SPA websites, but for brochure sites, it’s always my go-to.

What would you say it’s good for?

Anything you might use Next.js for but without the pain and frustration of Next.js. It’s ideal for sites needing SSR/SSG for SEO, content-heavy sites like blogs or documentation, and e-commerce.

Astro is particularly beneficial when you want to maximize page performance by minimizing JavaScript payload, allowing for limited interactivity with small islands of interactivity.

You can easily mix and match different paradigms (vanilla, React, Vue, etc.) for specific needs. For example, imagine a product detail page making an API call to an edge function to retrieve availability based on the user’s IP location or user-entered Zip.

Overall, Astro.js really excels in use cases where you can SSG your content, distribute it via CDN, and use limited JS for targeted interactivity.

Have you used it before?

Yes, our team switched from Next.js 12 to Astro. We found issues with Next.js 12 regarding 3rd party libraries and unexpected renders. Next.js 13 felt unstable and didn’t solve our problems, so we opted for Astro.

We experience less jank with Astro. The clean HTML output significantly improves our Google Search performance, reflected in Lighthouse scores. Although we still have to optimize for speed, we find Astro easier to work with compared to Next.js.

Will employers be looking for Astro developers anytime soon?

Unfortunately, it’s unlikely. Many large enterprises and growing startups are invested in Vercel’s Next.js hype. Companies like Target and Walmart use Next.js, so there are fewer large employers adopting Astro.js.

@Brady
We use Next for our corporate website (I’m on the design team) and I find it slow and janky.

Ash said:
@Brady
We use Next for our corporate website (I’m on the design team) and I find it slow and janky.

That’s not Next.js’ fault.

Ellis said:

Ash said:
@Brady
We use Next for our corporate website (I’m on the design team) and I find it slow and janky.

That’s not Next.js’ fault.

Well, it’s not NOT Next.js’ fault.

@Brady
That’s why I’d never leave React or Next.js in a corporate environment. At least you can Google solutions for any issues.

@Brady
Astro is cool, but it falls short compared to Next.js. I’ve led frontend projects with both frameworks. Astro works great for multi-page sites, not so much for SPA-style apps.

Remy said:
@Brady
Astro is cool, but it falls short compared to Next.js. I’ve led frontend projects with both frameworks. Astro works great for multi-page sites, not so much for SPA-style apps.

> Astro is really cool but it falls short of what you can do with Next.js.

Like what?

Astro is pretty great if you’re mostly doing websites that need a little backend and well-contained JavaScript. This is the case for most web apps, actually.

If you’re using Astro, you’re applying the same skills as with other JS frameworks, so don’t tunnel-vision your job prospects. Astro is quite efficient for SSG. Personally, I’m not a fan of the template system and JSX inspiration but it gets the job done. It might fit for developers/projects or even Publi for users who aren’t hardcore developers but can manage a website.

It’s great! Even nano stores for state management are phenomenal.

I’m currently building a large corporate web app in Astro. I decided against Next.js since the app must be self-hosted, and Vercel is unresponsive about supporting non-Vercel infrastructure for their edge runtime. Astro fits better for portability and avoiding vendor lock-in.

Astro’s smaller surface area makes onboarding less experienced team members easier than with other frameworks. You can completely skip client-side stuff. I believe Astro projects will age gracefully since it adheres closely to HTML/web standards that shouldn’t change in the next decade. Other frameworks often hide details behind abstractions that may not be consistent in five years, requiring more maintenance.

Overall, I’m very happy with the developer experience in Astro. I miss a few things (snippets, layouts avoiding reloads), and Astro isn’t suitable for highly integrated interactive experiences (like photo editors). However, most web apps I’ve built can be done with Astro.

I believe it’s the best framework out there! I’ve used Next, Nuxt, and SvelteKit, but I really value the liberty, performance, developer experience, and capabilities of Astro.

It’s really good for basically everything.

Great for static content—I use it for sites that don’t require much JavaScript.

However, I’d never use it for a full-stack application, for example.

I’d rather stick to SvelteKit. I don’t care about native Svelte support. If I’m using Svelte, I’m only using SvelteKit.

Jem said:
I’d rather stick to SvelteKit. I don’t care about native Svelte support. If I’m using Svelte, I’m only using SvelteKit.

How do you like SvelteKit? Any pros/cons you’ve found? I’m noticing more devs, especially influencers, using it to build their sites, but I haven’t seen many big companies utilize it yet.

@Luca
Check out GitHub - janosh/awesome-sveltekit: Awesome examples of SvelteKit in the wild and https://madewithsvelte.com.

@Luca
My point is that Astro feels like a middleman. You can use any framework with it to supposedly achieve faster sites. I prefer using SvelteKit since I’m knowledgeable enough to make it performant.

As a freelancer, I enjoy using SvelteKit. I used to build sites for clients without frameworks, just plain HTML, CSS, & JavaScript, and SvelteKit provides a similar feel but with extra benefits. If you want to work in web development, stick to React.