Honestly, I didn't expect the release to generate much conversation at all. There's a lot less focus/interest in front-end frameworks generally these days, for obvious reasons, so I figured the median reaction would be 'oh, the Svelte team are still shipping? good for them' and not much more. It's a joy to see the conversation here whether it's from happy users, or people who prefer different things, or people who have fully outsourced the writing-the-code bit to LLMs and straight up don't care about the underlying tools any more.
Which I totally understand! As framework authors we're very persnickety about the details of the code, so we've been on the slower end of the LLM adoption curve, but even we're starting to delegate more of the work to agents. If you're prompting an app into existence it's natural not to care that much about the particulars. But a question I've seen several times here, and in the zeitgeist more generally — 'in a world of agents why should I care about Svelte vs React vs whatever?' — deserves an answer.
And it's this: your app will be _better_ if you use Svelte. Your JavaScript bundle will use fewer bytes, your server-side rendering will take fewer milliseconds, and your users will reap the benefits: faster, more efficient apps. For all that the agentic revolution has created a huge _quantity_ of software, the _quality_ of the average app stubbornly refuses to budge. If anything, software is becoming less reliable. I think we've all felt that. Using a framework that treats things like accessibility and progressive enhancement as core concerns will help you steer towards better outcomes. (Your agent will use fewer tokens as well.)
Of course, this was always the pitch! Agents don't change that. And that's why we're still shipping and still sweating the details. We have some stuff coming up that we're unreasonably excited about — we really care about this stuff, and we're very grateful to be part of a community who shares that passion. Thank you for the support, it means the world to us.