100 Days of Node.js vs Bun - Who is the winner?

Was useful

Good content. I believe BUN will get better in over time tho.

Been using Bun in production for running scripts written in typescript for deploying Remotion apps to Lambda. ts-node would never work and Bun does work. I understand not everyone can switch to it, but using it for isolated purposes like mine makes sense IMO :D

What about tsup? what are your views on it for just running things locally?

great review and examination after actually using it in a business. We tried to add Tailwind to our aging Angular project and even that took more than 2 days and we gave up. Unfortunately with agile you can't stick to these changes and be productive at the same time.

nay, the hype was worse than nfts

I honestly think you're the "hype explorer" , thank you for the time for your advice. Just good work

ts-node also works na?

Thanks for trying it out and sharing your experience! To me it feels like bun gets better every day. I hope that you'll give it another chance again in a few months when it had even more time to mature.

Lets say I want to learn backend Javascript what do i start with Bun or NodeJs. Does Bun have all things Node has Like Stream Buffer etc.

I HAD to subscribe after this video, thanks for giving a perspective on a real company thought proccess instead of making a hiporbole statement and putting a scared face on the thumbnail

Don't be too hard on the Bun, every startup project is bad at first, but with time it will get better

Did you do something like this for deno

i tried bun for a while too, it has some issues.
but i like how simple and straight forward and clean the api is.
also fell in love with the macros, literally removes the need for preprocessing for most stuff related to code generation.
i really think about using it more often after they fix the bugs

I think they need to dial back on the borderline false advertising. It's not necessarily false advertising but it's very close to the line imo. Claims of "drop-in replacement" for Node when they know they aren't is not a great look imo. I have respect for what Jared and the team does and in the time they do it, but they just have to stop.

If Bun truly want to be seen as compatible, they should do what Deno does and run campaigns for people to replace Node with Bun as a challenge to help improve the compat. Of course they'd need to incentivize people to do it.

what bun version are you using right now

I’m sorry you ran into these reliability issues. Our focus today is on making Bun more reliable (and Windows). It’s going to get a lot better over the next few months.

man someone needs to give you and your team some award for overEngineering

I am using bun in a couple of project in production. have no issues with it, in terms of development, ci/cd, or runtime. I have use it with Elysia, Svelte and Astro. But all of them are new project from scratch. Tried to migrate few past project to bun, but the time it will take to make it work is really not worth it. At least for me, all my future project will start with bun.

NX workspaces mate, Bun and Deno support in NX workspace moneorepos.

You have an accent… you sound like a CEO 😅😅😅😅 I bet we can all agree that there are a lot (of them in the top 500 company) but I wanted to make a compliment and I was thinking about Microsoft CEO

Keep up your good work its appreciated 🎉🎉🎉🎉

Not sure how much you know about bunjs but let's just clear one thing here that bun doesn't rely on nodejs crypto it's just trying to have the same functions and features implemented exactly (syntax wise) so old nodejs packages can work without building them again. That's how Bun will become the drop in nodejs replacement.

how is the memory footprint compared to node js

Just hype

Ugh why do JS devs keep reinventing the wheel. Anyone remember Deno? What happened to that?

In which framework your Back-end is ?

Nice job keeping it real. Bun needs a lot of bake time before if can can fully live up to the hype/promise. Nothing against the great work the bun team is doing (thank you!) -- any such solution would require the same. As a runtime, I ran into the issues you're describing from onset and quickly established not ready. However, Bun has overall been working out well for me as a drop-in pnpm replacement (not using mono-repo), as a test runner, and it's great for running ad-hoc ts scrips from cmd line (tsx works well too, but it has annoying start-up latency for me that's absent in bun). There's been a few times I considered reverting back to pnpm/tsx/jest stack, but I'm holding out and sticking w/ bun because the dev teams seems hungry, and I like the direction.

Thank you for making this. I am a SaaS founder and I was checking out if it is a good idea to use Bun as the REST-Backend with MySQL to handle the management console of our product.


Now I know node express is the safest option even if it is slower.

It's really hard to run Bun on Windows
I tried WSL but it keeps crashing

Why to create new framework. Improve the existing one

Thank you for your honest review and recommendation. I hope and am convinced that Bun will get there, but it's not ready yet. I also think that for node pnpm, turborepo and tsx are solid and fast tooling solutions.

I tried it multiple times & got frustrated each time & gave up (weeks ago last time) & it’s really buggy. I wouldn’t really call it release quality. I’m back to pnpm. Try to debug with it with vscode. What a pain!!!!

I agree with most what you said.
But I really enjoy Bun, and I also keep Node LTS around, just in case.

This is a very helpful review. A lot of people, including me, delve into a new area and are confronted with several questions: Am I looking at the latest docs and Am I using the latest versions of the tools. You'll jump into a new tech area and see that several recommended items are in 'deprecated' or not-the-latest-version based on the tutorial or even GPT you are looking at / conversing with. So we are always looking for the latest just for code health, but then we also hear promotions of "The Newest and Improvest" novel tools. It can be difficult for an experienced technologist, let alone for newbists.

"It doesn't matter how fast it is if it doesn't work"

over all, for enterprise product like as goverment apps or retail corporate, are BUN js consumpt, are suitable for now ? or needs several time for waiting BUN be perfect technology ( framework ) ? thanks bro

Deno is the winner

Node evolved ober years, how do you expect bun to be perfect after 100 days. If people do not start using it, it will also never get to where node is.

Reminds me once again how screwed up the js ecosystem is