Faster initial loads with Response Streaming 🏞️ + SvelteKit 🔴 LIVE Coding & Chill

Let’s figure out how to start serving our apps faster, by streaming responses after our initial HTML load! My latest "edited" / streamlined video was a walkthrough on making this ChatGPT / OpenAI powered app: 🚀 Check out the app yourself! https://hellothere.cat/ 🪞 Make sure to try out your name, like https://hellothere.cat/Johnny 💻 See the code, open-source and free: https://github.com/jmagrippis/hello-there YOU TOO can leverage generative AI, just by watching that video 😄 The one thing we didn't really cover, was mitigating the fact that GPT4 is quite slow to deliver a full response. So, today, we'll investigate how we can start showing our app to users ASAP, thanks to Readable Streams! ⚡️ — COMMUNITY — New DISCORD SERVER for us 😱: https://discord.gg/eR5Q52Sfm3 — 🥳🥳🥳 — — Agenda — [ ] What are we doing today? “Hello There” app? ChatGPT with SvelteKit video?! https://www.youtube.com/watch?v=Uw5GZg96kD8 [ ] Streaming with promises: https://kit.svelte.dev/docs/load#streaming-with-promises [ ] Gotchas & deploying on the EDGE and other platforms which support streaming [ ] What if responses are sometimes slow, sometimes fast? 🤔 React Suspense equivalent? [ ] Can we load data that’s ITSELF a stream?! https://platform.openai.com/docs/api-reference/chat/create#chat/create-stream [ ] Svelte Hackathon & this cool submission for stream overlays!: https://github.com/edsunman/overlay ——⬇️ BONUS ⬇️—— [ ] Typing animation [ ] NEXT STEPS: Some of the “BONUS”s we didn’t get to? Let me know! — HAVE FUN — No affiliations whatsoever: if I show something, you know you're hearing my unfiltered thoughts 😄 — Highlighted software — My own website: https://magrippis.com/ Search for `@jmagrippis` to find me on socials like Instagram & Twitter! And remember our DISCORD SERVER 😉: https://discord.gg/eR5Q52Sfm3 Between the comment section and the socials, let me know **somewhere** what you’d like me to cover next 🙌