Back to stories
typebase.io

Show HN: Typebase – A single-folder back end you write in TypeScript

118points byandrewww-dev5d
Hey HN!

I built Typebase, a library that gives you Convex's DX with Supabase's openness.

After trying Supabase I liked how fast it is to spin up a DB and auth, but really didn't like using RLS and SQL for authorization. With Convex I loved how your server "lives" in your code, but disliked the DB model and the realtime-first defaults.

With Typebase you just write TS files inside a typebase/ folder in your existing repo. You can define your DB tables inside a schema.ts file and export server functions that your frontend calls like local functions, fully typed. Auth is built in.

Then one CLI command uploads your server to any of the available providers (Vercel, Cloudflare Workers or Deno Deploy for the servera and Neon for the DB), or generates the code so you can deploy it wherever you want.

Built on top of oRPC, Drizzle, and better-auth.

Happy to answer any questions or feedback!

Comments

34 total · 34 loaded

Ranked by HN
It’s like next.js but for backend.

Personally not for me. These are very productive once you learn the shape they expect.

But the wheel needs to be reinvented on how to do everything.

I ran have straight forward verbose code with no hidden control flow

Am I missing something? The entire point of next.js is that it's backend+frontend in one project
apsurd2don HN
that’s likely the promise but it’s so confusing and high cognitive overhead to maintain that backend <-> frontend relationship. ‘use client’ - i think the special sauce for why people use nextjs is all frontend based. the frontend that comes from the backend.

nextjs dx is a special kind of hell. disclaimer: im just a simple rails guy.

chrysoprace2don HN
Next.js is a poor choice for a client-side application. It is first and foremost a server-side framework, and trying to force it to be client-side is to work against it. There are so many better options than Next.js for dedicated client-side front-ends.
Nextjs is also a bundler for web applications. And it deploys the front end assets.

It's yet another backend framework, added that it only runs on vercel infrastructure.

There are better front end and back end servers, those that simply rely on nodejs to serve. Next.js does that, without vercel.

I don't understand this perspective. I've developed, delivered and maintained countless Next.js projects and never come close to considering it "high cognitive overhead." It's among the nicest DX I've experienced. It has a lot of features, I guess, but you don't have to use them if you don't like them. Next.js has very little to do with the frontend. The frontend is literally just react, which is why I commented in the first place. People keep acting like it's a "frontend framework" which doesn't make any sense at all.
React is a frontend framework. How do you get around that? It's invented entirely to handle client-side UX and reactive DOM updates which have to be DOM because it for the browser which is a frontend. Are you saying nextjs is the backend and react is just react, it's not nextjs' fault?

I'd say we disagree on the simplicity then: hydration, nextjs API reinventions of every interaction else 'use client' as an escape hatch which means it runs on the client but then what does the non 'use client' code do if nextjs is not a client-facing framework?

NoItsDanger2don HN
I’ve been saying for years we need CCR (Client Centric Rendering).

What we do is have the server just serve the page and let the client handle most of the rendering.

It’s novel and innovative

I let the LLM tokens flow. Usually it ends up saving me innovation tokens.
To be honest, the use of AI in your marketing doesn't bother me. But paired with the fact that you have a newish HN account with no comments ever, and no comments on your own "Show HN" in response to others, doesn't put off strong trust signals.
The pronunciation of "Vercel" near the end of the "Typebase in 100s" video was VER-cel, with the emphasis on the first syllable, instead of ver-CEL with the emphasis on the second syllable.

Was the video generated by AI?

I have deployed and maintained plenty of diff setups on GCP, AWS, colo, and I have also used AI for coding, my concern is do we really trust AI to do prod deploy and maintenance of the backend yet? Not to mention what are the limitations here, someone already mentioned queues? But for a straightforward site with limited complexities, it might be cool to try out.
phren0logy2don HN
I love the (well-disclosed) fake testimonials. Hilarious!
> “I’ve told four separate therapists about Typebase. Two stopped taking me as a client. The other two are now shipping an app with it.”

So good. I laughed out loud at this.

Scratching my head a little bit here, I'm not sure how this differs from NextJS (or Vite) with a layer of drizzle & better-auth.

Not that that layer isn't important. We ended up building something like this at Ingram, nextkit: https://github.com/ingram-technologies/nextkit - but we distribute it as a set of micro libraries that work well together rather than a unified framework. (Auth is nk-auth, built on top of betterauth, and sql is nk-db, built on top of drizzle)

You got to pretty much the same conclusions we did though regarding the "why".

rgbrgb2don HN
I’ve been curious about trying the Supabase type of RLS model for an app but haven’t yet. I like the idea of the db defined api like postgraphile and hasura but had trouble with testing and ergonomics. This looks pretty nice though, it’s cool that you’re thinking about ergonomics for working with coding agents. I feel like a hard thing with that rn is that everyone is kind of figuring out different workflows for AI coding right now so it’s hard to optimize DX. Can you talk a little about your personal agent workflow and how this fits it?
k4rli2don HN
I've only had experience in one full-stack project with supabase RLS so experience is limited. However it seemed like a mess. Seems much more human-maintainable to have a classic spring/golang backend. For sure in some projects, esp mobile apps, it makes more sense, and fable-level models might also keep it from being full spaghetti.

Also Supabase at that time wasn't reliable and it's not really made for self-hosting. Felt almost as bad as Github stability these days.

Still, based on this limited experience, I've been avoiding RLS and supabase.

I thought supabase was mainly for self hosting? Like it was self-hostable firebase... Not too familiar with it though.
Sounds very interesting. I have to agree I didn't like RLS as well. I mean, it's useful but it seems too many knobs to turn...
I was about the ask a question but you just answered how it is built. I guess I will try it out, it sounds really good.

I love your approach, fake testemonials and all.

ramon1562don HN
Fake reviews are a hard no-no imo. Really weird addition, but I suppose it wasn't you who got the idea.

Looks an awful lot like a orpc, just built on top of drizzle instead of raw queries. I'm sure some people want this, but it seems like gross duct tape that will bite you in three years

dinkleberg2don HN
I’m very surprised to see others commenting here that they like the fake reviews. Claude loves adding that shit, but you’re allowed to modify and remove the junk it produces.

> * None of these people exist. We checked. Twice. Legal is chill

Maybe I’m just the old man yelling at the clouds, but can we not have some integrity and actually put the truth out there? There is nothing wrong with using AI, but we’re still responsible for our work.

I don't know, I have plenty of gray hair, but I still love a little tongue in cheek humor. I still remember the first cold email I got asking for 12 minutes of my time - I replied. Now if the disclaimer was buried at the bottom of the page and this product was asking for money, I would agree there is more of a trust issue here.
Very cool!

Note that you can run Convex on top of your own Postgres, and we do this for our internal apps, with very nice performance (plenty for our internal use cases).

very cool. though the deploy wasnt clear if i can just run migration on some pg url. and it wasnt clear if i can bind some middleware into my existing app server (like express)
What, in your opinion, is this not a good fit for?
Enterprise software, stuff that you expect to scale exactly because it has queues and distributed computing.
SifatAhmed2don HN
This looks great. Question specifically on the Cloudflare Workers target: when you deploy there, does Typebase use D1 for the DB, or does the Postgres/Neon backend stay the same regardless of which server target you pick (Vercel/Workers/Deno) - i.e. is the DB layer decoupled from the server runtime, or does choosing Workers also mean choosing D1? I ask because I've shipped a small Workers+D1 service and the main friction I've hit isn't the server code, it's Workers' constraints around D1 (no long-running connections, HTTP-based driver quirks) - curious whether Typebase abstracts that away or whether you still feel it at the DB layer when targeting Workers specifically.