pothos VS vite

A project I maintain Pothos uses a global namespace with a bunch of interfaces to allow plugins to extend interfaces defined in core or other plugins. This allows plugins to add new options and methods to objects and classes without the other packages needing to know anything about them.

I noticed this being shared around on Twitter the other day - pretty handy, as I'm currently trying to architect a similar experience for my job using Pathos and graphql-codegen.

- tRPC

But I'd likely throw out Clerk a cheaper option:

- Supertokens, and also since Supertokens is easy (lots of enthusiastic reports about it), has a managed solution (which is cheaper than the alternatives), is secure and scalable (rotating refresh tokens with JWTs), open source, has magic links, and the architecture of Supertokens would allow me to simply and quickly eject to self-hosting it if/when I'd eventually need to (if the app ever reaches mass-market scale).

And I might throw out tRPC for the equivalent GraphQL experience (esp. if business strategy dictates I need a 3rd party API):

- GQty.dev on the client, for inferred queries/mutations. For rapid dev speed. Simple code example: https://gqty.dev/docs/intro Then move to URQL or Relay at scale, or just skip GQty and go with URQL from the start (if scalability trumps dev speed).

- Pothos http://pothos-graphql.dev on the server, for auto building the schema from your TS code (aka. code-first). Better than Nexus (e.g. Max Stoiber moved from Nexus to Pothos on his Bedrock starter template because Pothos is best in class: https://bedrock.mxstbr.com/tools/pothos/ ).

And I might throw out NextJS (Webpack) for the equivalent experience in Vite:

- vite-plugin-ssr, since both architectural control (libraries > frameworks) and Vite rocks. I'd likely then have to make solito-vite https://github.com/nandorojo/solito/discussions/157 to have a unified navigation between React Native and Web, but Solito is allegely tiny, so recreating it should be doable.

(If doing all of these replacements, maybe starting from scratch would be easier than modifying create-universal-app ... That said, I think if someone made a starter repo with the above choices it would be a real killer!)

Then I'd also likely use:

- Vercel (and try their Edge Functions, for a serverless sweet v8 isolates experience without slow cold starts), or maybe Cloudflare Workers (cheaper, slightly more hassle?) for hosting.

- Planetscale or Supabase for the DB. (Not brave enough to try EdgeDB or SurrealDB just yet, though EdgeDB is close..) Unless I had a specific use case where a more specialized/optimized DB would make sense.

This stack should stick even post-MVP, as it's not only optimized for a solo developer but for scalability.

Have you used Pothos? It's a way to make GraphQL schemas in TypeScript, in a type-safe way. So the creator of Prisma Client Rust is thinking about making a Pothos-style API based on the t builder pattern:

I would recommend Pothos (https://pothos-graphql.dev/) as a more modern alternative to typegraphql or nexus.

Depends on language, I've build GraphQL servers in a few, though mostly JavaScript and Python. For Python I used to use Graphene, these days I use Strawberry.

For JavaScript, I originally used graphql-js and express-graphql, as these were the original libraries and I was a literal day 1 adopter. All the libraries are essentially just wrappers around graphql-js, so it's still viable to use directly. But for schema-building I now use Pothos (https://pothos-graphql.dev/), I'd probably use graphql-helix as the http layer (https://github.com/contra/graphql-helix).

Pothos is a breeze of fresh air when it comes to building GraphQL APIs. It is a library that lets you write code-first GraphQL APIs with an emphasis on pluggability and type safety. And it has an awesome Prisma integration! (I am genuinely excited about this one, it makes my life so much easier.)

In today's article we are going to create a GraphQL api using the Koa framework together with the GraphQL Yoga library and Pothos. In addition, we will use Kysely, which is a query builder entirely written in TypeScript.

If you're a GraphQL developer, Pothos is the best example - all your user-defined types just fits in it like a glove 99% of the time. It definitely makes the most use of TS generics.

https://pothos-graphql.dev/

(I'm a bit sleepy, so this is the main one I can think of at the moment that I really enjoy using.)

Given our team's collective proficiency within the React ecosystem, we decided to leverage this expertise for our project. Initially, we contemplated utilizing Next.js; however, due to the limited practical experience with this technology among key engineers and the pressing timeline to develop the first prototype, we opted for a Single Page Application(SPA) approach. For bundling, we selected Vite, primarily due to its super fast build times, simplicity of configuration, and potential for a nearly seamless transition to server-side rendering.

We are continuing to add new project templates for various types of projects, and we've recently created one for the infamous combination of React with Vite tooling.

Vite focuses on providing an extremely fast development server and workflow speed in web development. It uses its own ES module imports during development, speeding up the startup time.

Vite and Next.js are both top 5 modern development framework right now. They are both great depending on your use case so we’ll discuss 4 areas: Architecture, main features, developer experience and production readiness. After learning about these we’ll have a better idea of which one is best for your project.

import { defineConfig } from 'vite' import react from '@vitejs/plugin-react-swc' import path from 'path' // https://vitejs.dev/config/ export default defineConfig({ plugins: [react()], server: { port: 3000 }, css: { devSourcemap: true }, resolve: { alias: { '~': path.resolve(__dirname, './src') } } })

I am currently utilizing Vite:

Homepage: https://vitejs.dev/

Without any adding any dependencies you can connect react props to raw css at runtime with nothing but css variables (aka "custom properties"). If you add CSS modules on top you don't have to worry about affecting the global scope so components created in this way can be truly modular and transferrable. I use this with vite.

Little confused as to why it has vite in it‘s name, it seems unrelated to https://vitejs.dev/

it says in their docs that they recommend Vite https://vitejs.dev/

it goes like this.

1. you create a repo folder, you cd into it.

2. you create a client template using vite which can be plain typescript, or uses frameworks such as react or vue, at https://vitejs.dev/guide/

3. you cd in that client directory, you npm install, then you npm run dev, it should show you that it works at localhost:5173

4. you follow the instructions on your url, you do npm install @web3modal/wagmi @wagmi/core @wagmi/connectors viem

5. you follow the further instructions.

> It seems like this is for npm or yarn to pull from a remote repository maintained by @wagmi for instance. But then what?

you install the wagmi modules, then you import them in your js code, those code can run upon being loaded or upon user actions such as button clicks

> Do I just symlink to the node_modules directory somehow? Use browserify? Or these days I'd use webpack or whatever the cool kids are using these days?

no need for those. browserify is old school way of transpiling commonjs modules into browser-compatible modules. webpack is similar. vite replaces both webpack and browserify. vite also uses esbuild and swc under the hood which replaces babel.

> I totally get how node package management works ... for NODE. But all these client-side JS projects these days have docs that are clearly for the client-side but the ES2015 module examples they show seem to leave out all instructions for how to actually get the files there, as if it's obvious.

pretty much similar actually. except on client-side, you have src and dist folders. when you run "npm run build" vite will compile the src dir into dist dir. the outputs are the static files that you can serve with any http server such as npx serve, or caddy, or anything really.

> What gives? And finally, what exactly does "browserify" do these days, since I think Node supports both ES modules and and CJS modules? I also see sometimes UMD universal modules

vite supports both ecmascript modules and commonjs modules. but these days you'll just want to stick with ecmascript which makes your code consistently use import and export syntax, and you get the extra benefit of it working well with your vscode intellisense.

> In short, I'm a bit confused how to use package management properly with browsers in 2024: https://modern-web.dev/guides/going-buildless/es-modules/

if people want plain js there is unpkg.com and esm.sh way, but the vite route is the best for you as it's recommended and tested by the providers of your modules.

> And finally, if you answer this, can you spare a word about typescript? Do we still need to use Babel and Webpack together to transpile it to JS, and minify and tree-shake, or what?

I recommend typescript, as it gives you better type-safety and better intellisense, but it really depends. If you're new to it, it can slow you down at first. But as your project grows you'll eventually see the value of it. In vite there are options to scaffold your project in pure js or ts.