'use client'; 💀💀💀💀💀💀 JSX loops: return ( {collection.map(function (item, index) { return ( {index}) {' '} {item.label} ) }) ); 🤮🤮🤮🤮 Hooking asynchronous code on mount: useEffect(() => { fetchApiResource(identifier) .then(res => { const {data} = res const sanitized = sanitize(data) setDataState(sanitized) }) .catch(e => setErrors(e.map(item => item.toString()))) .finally(() => setLoading(false)); }, []); 🤡🤡🤡 CSS-in-JS (because not React team's problem?): {children} 🧑🦽🧑🦽🧑🦽 Ecosystem of poorly designed, slow, boilerplate heavy, bad DX, un-testable libraries that are supposed to be included with the framework itself: Redux (🥱💤): Flux in 2024? Zustand (👨🔧👉💥): Impossible to test. Formik (📃 🔌): Boilerplate unreadable, even more when using schema validation. StyledComponent & Friends (🧑🦯): String... Intellisense? Code formatting? Linting? Framer Motion (🐌🫨🩹): Heavy and slow. Who thought it was a good idea to describe animations inside HTML?
I can't believe a PHP framework has been using Vite for over 2 years now while Next, which is one of the most popular JS frameworks today, is still stuck with something that's quite similar to Webpack
Awesome presentation! The vision here is great and it made me so excited for the next year. Can’t wait to try it out as soon as it goes out, though Vite is currently very fast in my opinion, as you mentioned that the majority of apps are small in size.
@@Alex1611AD He didn't. 5 million is nothing for a project like this. If I make a colorful PowerPoint with some AI idea and reach out to some investors, I have 30 million a week later (and I am not talking about something crazy). Companies really need to start funding these projects.
Talking about the "massive inefficiencies in the build pipeline" in vite when we shaved off 95% of the build time of our previous projects compared to vue-cli, is saying something 😂
tbh sending individual ESM modules during development currently makes vite unusable for majority of corporate projects, i tried running ours and load times were over 40s with thousands of requests. Looking forward to bundle mode to give it a try again.
Vue must rely on dinosaurs like Laravel, Firebase and RoR clones... PLEASE Evan build your own websocket first API framework on TS. One where you're subscribed to changes by default instead of constantly polling REST calls.
@@RobertBMenke I'm just annoyed that after 10 years, Firebase is still unique. FeathersJs was close, but v5 is unusably complicated, and v6 is looking like a closed source turn-key product... Just like Firebase.
I don't know if I understand your problem correctly, but there is a handful of self-hosted Open-Source alternatives to Firebase (Supabase, Pocketbase,...)
@@alamosdark4517 Supabase is SQL based, Firebase is document storage backed, like MongoDB. Never heard of pocketbase until now... It's not stable and it's SQLite backed (massive problem for larger apps), but other than that it's looking great.
@@alamosdark4517 Supabase is Postgres backed and uses SQL-like syntax and is therefore terrible are representing hierarchical data. Unlike Firebase. Never heard of Pocketbase until now... Looking into it. It's backed by SQLite but comes with its own schemas and Typicode, the creator of LowDB is following them... Might flourish into something good. Thank you!
Depends on what you're doing. Not every project is best served by a no build setup, and it means you need to write *everything* yourself. If you're writing a "web page", yes, no build is probably a good approach; if you're building a "web app", probably not.
All this nonsense is the never ending battle to fix JavaScript problems using more JavaScript. It will continue to grow until dissolution. The immense fragmentation and the underlying language (even with typescript) is driving me away. I have more interesting things to do with my life than to learn yet another JS tool to do a job that I already know how to do. This people are profiting from your time.
@@sgsniper1 I do it all the time, but not at my day job, I have no decision power there, I also don't care at all about what they do tech-wise, we just sell stuff online, that's what matters.
Excited to see the yearlong vision coming to life!
I think this is the best shot on a unified toolchain that the JS ecosystem has for years
i still believe, that having a no build step is the way to go.
Next.js really needs to adopt Vite, it is getting ridiculous
totally agree with you 💯
Perhaps it's time to move on to Remix
unlikely to happen any time soon, but I agree. nexts builds are soooooooo slow man
'use client';
💀💀💀💀💀💀
JSX loops:
return (
{collection.map(function (item, index) {
return (
{index})
{' '}
{item.label}
)
})
);
🤮🤮🤮🤮
Hooking asynchronous code on mount:
useEffect(() => {
fetchApiResource(identifier)
.then(res => {
const {data} = res
const sanitized = sanitize(data)
setDataState(sanitized)
})
.catch(e => setErrors(e.map(item => item.toString())))
.finally(() => setLoading(false));
}, []);
🤡🤡🤡
CSS-in-JS (because not React team's problem?):
{children}
🧑🦽🧑🦽🧑🦽
Ecosystem of poorly designed, slow, boilerplate heavy, bad DX, un-testable libraries that are supposed to be included with the framework itself:
Redux (🥱💤): Flux in 2024?
Zustand (👨🔧👉💥): Impossible to test.
Formik (📃 🔌): Boilerplate unreadable, even more when using schema validation.
StyledComponent & Friends (🧑🦯): String... Intellisense? Code formatting? Linting?
Framer Motion (🐌🫨🩹): Heavy and slow. Who thought it was a good idea to describe animations inside HTML?
Vercel's control of Next.JS needs to be stopped.
What a great communicator.
I can't believe a PHP framework has been using Vite for over 2 years now while Next, which is one of the most popular JS frameworks today, is still stuck with something that's quite similar to Webpack
not even similar, it's still using webpack for prod builds
Microsoft should rewrite TypeScript typechecker to Rust
🦀
I really want to try the formatter and linter, eslint and prettier are really slow
Try Boime while OXC is WIP
@@olegafx OXC's is mature and has more rules than Biome
@@zettca Linter - yes, formatter - no
Awesome presentation!
The vision here is great and it made me so excited for the next year.
Can’t wait to try it out as soon as it goes out, though Vite is currently very fast in my opinion, as you mentioned that the majority of apps are small in size.
Hi i Evan You - creator of all your stuff.
Only Next resist out of pure foolishness
shame on nextjs 😅😅😅
They’re busy knee deep collecting. 😂
Please find a way for big corps to start paying for the hard work of Open Source !
@@Alex1611AD He didn't. 5 million is nothing for a project like this. If I make a colorful PowerPoint with some AI idea and reach out to some investors, I have 30 million a week later (and I am not talking about something crazy). Companies really need to start funding these projects.
I hope everyone will use vue!
time to write Cuda optimized js tooling 😅
That is so cool
Talking about the "massive inefficiencies in the build pipeline" in vite when we shaved off 95% of the build time of our previous projects compared to vue-cli, is saying something 😂
JavaScript 2.0
tbh sending individual ESM modules during development currently makes vite unusable for majority of corporate projects, i tried running ours and load times were over 40s with thousands of requests. Looking forward to bundle mode to give it a try again.
Yea… exactly why Deno will reign.
why
Hobby $?
the only benefit i saw is you face your errors much quicker, am i right?
Vue must rely on dinosaurs like Laravel, Firebase and RoR clones... PLEASE Evan build your own websocket first API framework on TS. One where you're subscribed to changes by default instead of constantly polling REST calls.
wat
@@RobertBMenke I'm just annoyed that after 10 years, Firebase is still unique. FeathersJs was close, but v5 is unusably complicated, and v6 is looking like a closed source turn-key product... Just like Firebase.
I don't know if I understand your problem correctly, but there is a handful of self-hosted Open-Source alternatives to Firebase (Supabase, Pocketbase,...)
@@alamosdark4517 Supabase is SQL based, Firebase is document storage backed, like MongoDB. Never heard of pocketbase until now... It's not stable and it's SQLite backed (massive problem for larger apps), but other than that it's looking great.
@@alamosdark4517 Supabase is Postgres backed and uses SQL-like syntax and is therefore terrible are representing hierarchical data. Unlike Firebase. Never heard of Pocketbase until now... Looking into it. It's backed by SQLite but comes with its own schemas and Typicode, the creator of LowDB is following them... Might flourish into something good. Thank you!
Next year:
Introducing…RUST. Just use Rust, it’s less complex.
Don't get me wrong, Vite is great. But if people used JS the right way, there'd be no need to use any tools at all.
Depends on what you're doing. Not every project is best served by a no build setup, and it means you need to write *everything* yourself.
If you're writing a "web page", yes, no build is probably a good approach; if you're building a "web app", probably not.
Right way to use JS - not use JS at all? JS itself is one of the worst programming language out there.
All this nonsense is the never ending battle to fix JavaScript problems using more JavaScript. It will continue to grow until dissolution.
The immense fragmentation and the underlying language (even with typescript) is driving me away. I have more interesting things to do with my life than to learn yet another JS tool to do a job that I already know how to do.
This people are profiting from your time.
I truly think that all this crap needs to stop and just get rid of all "build" steps.
Use vanilla js you don’t have to build anything, but would you?
@@sgsniper1 I do it all the time, but not at my day job, I have no decision power there, I also don't care at all about what they do tech-wise, we just sell stuff online, that's what matters.
@@mvargasmoranthanks, God, you have no decision power on your job