I find it sad that Astro advertises itself this way, because I think that it is perfectly capable of building web projects of any complexity, simply by means of the component libraries you can plug in.

What makes it so great is not that it serves a particular niche (like "content-driven websites") but that it provides a developer experience that makes it incredibly easy to scale from a static website to something very complex and interaction-heavy without compromising UX.

That's the point. It's a war on recentish industry standard to develop all projects using same tools made by/for huge organizations working on huge projects.

Same thing happened with microservice architecture.

The number of web projects that fall into these categories vastly outnumber “complex” projects. For complex there is always WASM.

Yeah, this is my gut-feeling too! Because the alternative for "complex" being discussed is Next.js, but that doesn't really help you with "complex" applications, and you still have to bootstrap a lot of infrastructure yourself (with dependencies, or by yourself).

What do you mean by complex?

More than a blog or a simple static website.

Like what specifically? And why do you think Astro is not a good fit for it?

You could have React in one Astro island and SvelteKit in another. That's complex enough.