Jon Higger (He / Him) Posts

Jon Higger (He / Him)

•Created by Jon Higger (He / Him) on 10/4/2023 in #questions

Is anybody else in here super hyped on Elysia

Just putting out feelers for this. The second I tried out Elysia months ago my immediate thought was "Holy shit this is cool", and I feel like I haven't seen much hype around it here and am kind of surprised by that. To be clear, if you choose to keep reading this has almost nothing to do with the fact that: • Bun is cool bun

• Elysia is fast <:elysia_love:1140316901973708850> But instead, here's why I think it might be something worth paying attention to from this community, which I originally became a part of because of the focus on efficient full-stack workflow through create-t3: 1. The typesafety is <:chef_kiss:1101121096318713866> Right off the bat as a backend framework it's easily, and by far the best typescript experience I've had in a javascript backend. 2. You can generate a trpc-like client by injecting the typeof app into something called an edenTreaty The thing I like about this client way more than trpc so far, is with your backend just build some REST style endpoints, then you can build a client like this... [read more here https://elysiajs.com/plugins/eden/treaty.html], which to me reminds me a lot of the talk that @theo gave a while back about graphQL and how trpc is like a tradeoff that locks you into the typescript ecosystem. Well this does not do that., because even if you can't use an eden treaty, you still have a REST architecture.

const app = edenTreaty<App>('http://localhost:8080')
const { data: nendoroid, error } = app.mirror.post({
    id: 1895,
    name: 'Skadi'
})

const app = edenTreaty<App>('http://localhost:8080')
const { data: nendoroid, error } = app.mirror.post({
    id: 1895,
    name: 'Skadi'
})

3. Given the above points, it feels like this backend framework has the capability of filling in many pieces to the puzzle of "How do I get fullstack typesafety, without feeling like I don't have a proper backend", which to me is what NextJS, Remix, SolidStart, SvelteKit all feel like they fail at on some level to me personally. I think that it also has the potential to calm down a lot of the fears around "If I build this app with a full stack framework, will it be a PITA to decouple the code later on if I need to for some reason?" For example imagine if you had a backend that looked like this

const dogs = ["doomslayer", "zoey", "bowser"]
const app = new Elysia()
    .get("dogs", () => dogs)
    .post("dogs", ({body: dog}) => {
        dogs.push(dog)
    }, {body: T.String()})

const dogs = ["doomslayer", "zoey", "bowser"]
const app = new Elysia()
    .get("dogs", () => dogs)
    .post("dogs", ({body: dog}) => {
        dogs.push(dog)
    }, {body: T.String()})

I'll use remix as an example here just because I know it better, you could do something like

const loader = () => {
     return {
         dogs: (await eden.dogs.get()).data
     }
}

const page = () => {
    const { dogs } = await useLoaderData<typeof loader>()
    const revalidator = useRevalidator()

     return (
        <>
            <DogsList dogs={dogs} />
            <CreateDogForm createDogFn={(newDog) => {
                 // since this is just a thin wrapper around
                 // fetch it works perfect both server and client side
                  eden.dogs.post(newDog)
                      .then(() => revalidator.revalidate())
            }} />
        </>
     )

}

const loader = () => {
     return {
         dogs: (await eden.dogs.get()).data
     }
}

const page = () => {
    const { dogs } = await useLoaderData<typeof loader>()
    const revalidator = useRevalidator()

     return (
        <>
            <DogsList dogs={dogs} />
            <CreateDogForm createDogFn={(newDog) => {
                 // since this is just a thin wrapper around
                 // fetch it works perfect both server and client side
                  eden.dogs.post(newDog)
                      .then(() => revalidator.revalidate())
            }} />
        </>
     )

}

3 replies

TTCTheo's Typesafe Cult

•Created by Jon Higger (He / Him) on 6/14/2023 in #questions

Buying a new Mac, How important is the Chip or having >16GB ram

I saw theo's recent tweet about definitely not getting a mac pro and getting a mac studio instead, and it reminded me how bad my current personal computer is. I would like to upgrade to an m1 / m2 era computer but am not sure how hard to go on it. My needs are as follows: • Probably would prefer a laptop as I work outside all the time • Audio Production • Rarely / Occasionally Video Production • Lot's of coding, lately typescript + VSCode for me has been way slower than Logic Pro for example • Live Streaming + Making Music at the same time • Live Streaming + Coding at the same time I'm thinking I'm going to buy a refurbished laptop, but I'm trying to decide whether or not it's worth the extra grand to get a macbook pro. My constraints that I know for sure I want is • at least 2TB HD • at least 16gb ream but then after that I'm not sure how valuable the actual chip changing is, and if it will yield important performance gains for me. • More than 16GB ram • M2 vs M1 (any Max / Pro model included in that) Do any of y'all have good insight as to whether or not shelling out the extra money for more RAM or for a better chip than M2 or M1 Base model would be worth the extra grande that it would probably cost? I'm beginning to get the sense that with the silicon chips those upgrades might be overkill for the kind of work I do (No gaming for me).

29 replies

Gaming

Programming