If you're presenting Web 1.0 server-side apps as an alternative to JavaScript, you're trading accessibility for archivability. With a modern SPA, I can have the

If you're presenting Web 1.0 server-side apps as an alternative to JavaScript, you're trading accessibility for archivability. With a modern SPA, I can have the user download the app once and cache it in a service worker. All further requests can be made as minimal as possible using JSON or GraphQL requests. This matters, a lot, to users connecting over slow cell connections in metered bandwidth. I care a _lot_ more about making my sites accessible to users in developing countries than I do about y'all HN folks that refuse to run JavaScript on principle. Figure it out. Sites with JavaScript are somewhat more difficult to archive, sure. We've done a pretty good job though. I'd be curious to see if anyone here can actually present a site that doesn't archive in the Wayback Machine or archive.today. We've thankfully developed better tools for this sort of thing than curl, that are entirely capable of archiving script tags (yes, even from CDNs!) and pertinent async HTTP requests. Finally, server-side rendering obviates this problem entirely. Write your same React app, but also have the server return some coherent HTML content for `curl`. Next.js and others do this automatically. Turns out the solution was more JavaScript the whole time.