they handled this with 4 1x dynos on Heroku (before switching to AnyCable they had 20 2x dynos for ActionCable).

(load functions call handle directly, there's no intermediate network requests.)

When fetch runs on the server, the resulting response will be serialized and inlined into the rendered HTML. This allows the subsequent client-side load to access identical data immediately without an additional network request.

I want to avoid nginx overhead (especially if they have tons of alias and rewrites) for in-server communication. Basically, you can have sveltekit server, backend server and nginx server, in that case, communicate inside your internal network will be very expensive like: browser->nginx server(10.0.0.1)->sveltekit server(10.0.0.3)->nginx server(10.0.0.1)->backend server(10.0.0.2) instead just: browser->nginx server(10.0.0.1)->sveltekit server(10.0.0.3)->backend server(10.0.0.2)

Node doesn't have a DOM available. So in order to render HTML we use string concatenation instead. This has the fun benefit of being quite efficient, which in turn means it's great for server rendering!

But is overhead always bad? I believe no — otherwise Svelte maintainers would have to write their compiler in Rust or C, because garbage collector is a single biggest overhead of JavaScript.

Impressively, altering state in Vue is not only more succinct, but its re-rendering system is actually faster and more efficient than React’s.

If you're using webpack with svelte-loader, make sure that you add "svelte" to resolve.mainFields in your webpack config. This ensures that webpack imports the uncompiled component (src/index.html) rather than the compiled version (index.mjs) — this is more efficient.

you may specify only the form state that you care about for rendering your gorgeous UI. You can think of it a little like GraphQL's feature of only fetching the data your component needs to render, and nothing else.

If those comments are loaded outside of the blog_post association, then attempting to reference the blog_post association from within each comment will result in N blog_posts table queries even if they all belong to the same BlogPost!