Never outsource your auth no matter what unless you're vibe coding and it doesn't matter. If that's the case, don't put auth in at all because it's just you anyway or just use vibe coded password lookups from database at that point.
Lastly - here's the law of mother nature:
Software funded by VCs will milk you and will milk you dry. Not now, not tomorrow maybe decade or two later when the hot potato being passed around reaches to the last party holding the bag that now needs its money back with all that others had their 10x returns already taken from them by soldng them at an inflated price to the last party holding the bag.
Can someone more intelligent then me tell me why should I offload my postgres users table to some 3rd party provider? Like what is so hard about keeping that table in my VM on hetzner that I have to give it off to someone else? It's not payments, it's just a few fields of data
SSO, SAML, SCIM, OIDC, OAuth, 2FA, passwordless auth, verification tokens, etc etc, And, variations of each for wildly popular systems you’ll be expected to integrate with but don’t support the exact spec.
For a while at my company, half our support engineers time went to handling random SSO issues that came up in our home built auth system.
I must as intelligent as you because I also never understood why things like supabase even exist. I believe this shows how much front-end dev world is detached from how things can simple and secure by default.
Do you say the same about AWS RDS. Are you saying VMs is all you need and it is a doddle for anyone with FE only experience to set up, maintain and scale.
The only project where this was the case that I didn't hate it was at a former employer, and it gave the responsibility of securing users to Auth0 and minimized our PII and attack surface, since even the login page was not hosted or controlled by us. Worse case you somehow hacked our users and got some free entree reward they had, otherwise good luck trying to get very little data.
It allowed us to do SSO for small one-off marketing / campaign focused sites. I could give a specific login URL and it would always log you in if you were already logged on.
Don't you wanna level up your career to become an architect? You can draw a box, call it "User Management" and slap "Clerk" or some other SaaS on it, and assume it's managed for you. This allows you to shove whatever requirements you want in that magic blackbox as you feel "it doesn't bring value" for you to implement.
People are very scared of messing up authentication and getting hacked. They would rather offload that responsibility to a third party and not think about it.
Unfortunately this is a common premise and on surface its a good idea too to let a expert in particular domain handle it. Where it gets muddy is when this third party are themselves learners and just see this as a good business opportunity
Why pay someone to build a house? I’m sure you could do it yourself…but that doesn’t mean that is the best use of your time in all cases. The analogy is basic but apt; not everyone needs or wants to run (or create) every mechanism. I don’t do all of my own hosting either and it’s not because I couldn’t, it’s that it isn’t worthwhile in my cases.
To expand a bit more: if a business is faced with a choice to save some money by increasing risk, having people who’s job it isn’t managing and supposedly securing that information, or to have a third-party who job is literally to handle and worry about those things, who carries independent insurance, and who is on the hook if they lose customer data, and in exchange the business is simply taking the risk of associating with business that could do a poor job — which of those options sounds more appealing from a business sense? It’s a lot easier to blame someone else than earn back trust for your own major mistakes because you tried to write your own software to save a little money.
If you’re implying that people should __always__ roll their own services and never vendor out non-core parts, the security industry would love to learn where you work.
> Okay, so… what are those cases? I’m also curious.
If you're willing to make a third party SaaS's uptime the ceiling for your own org, you can delegate auth. Github might not be a good choice for SSO.
If you're not threatened by per-user-per-month fees, you can delegate auth.
If your threat model is compatible with a third party having visibility into your user's network location and the frequency and duration of their activities across your org, you can delegate auth. (Okta will probably not inform your competitor that your main sales guy is in North Carolina this week and has logged in from the conference room wifi of your competitor's main client.)
If you can trust the third party to not allow an interloper to bypass your requirements, you can delegate auth.
I am just as confused as you. My 2c: For a broad range of requirements, running your DB directly and managing auth with Django or similar is easier. Perhaps at enterprise scale, this changes.
That's what they did. They migrated to Better Auth, which stores everything in your DB. It's the equivalent of Django auth for the Typescript ecosystem.
The tl;dr of the article is that there are auth specific features that are not differentiated but that users expect. Just like you might outsource pieces of functionality like data storage and message sending to specialized servers/libraries/applications, you can do the same with authentication.
The article could use some improvements, tbh, it is 2.5 years old.
Social logins, email logins, password resets, multi-tenant, organizations, many to many users to organizations, etc etc. Not necessary for MVP, but can definitely be painful hacking in later if the MVP hits.
What you are talking about is in a large part authentication. You can do authentication using an external service and still have your user table locally. You can also do authorization locally with a local session table while leaving authentication to a SaaS.
Hey, Bereket from Better Auth here. I started Better Auth to solve this exact issue for myself, and it later turned into a company. It always give me joy to just see others getting the same value from it :) There is a lot to work on, would love to know what we can improve
I think auth is complicated outside of browsers too. But browsers do make some things uniquely confusing, especially cookies and general security primitives are full of footguns
I enjoyed the Supabase migration article from a while ago (https://blog.val.town/blog/migrating-from-supabase) as well. There's a shortage of good, honest writing on long-term engineering decisions, please keep up the blog!
> A hard lesson you learn building a complex system is that its reliability is the minimum of the combined reliability of its critical parts.
It's worse than that, the combined availability is the product of all components in the critical path. If your software, the authentication layer, and the cloud provider each have 99% availability, and any one of them can bring your service down, then your final availability is just 97%. With eleven components like that you have zero nines of availability.
That's why reducing components and going for reliable solutions is so important. I'm happy that the team took this path.
Learned this one the hard way during the last major CloudFlare outage. I don't use them, but their outage bricked my app for hours anyway because the Auth0 public keys used to verify JWTs were served behind CloudFlare, breaking the entire auth chain. Fun!
This is why I'm so thankful I went with Lucia early. They sort of sunset their library and replaced it with documentation (and some small utilities) for how to manage and host authentication for yourself. It's always presented as some big, scary thing you can't manage yourself, but I found that taking the week to learn how security and basic salting works, I was able to feel more confident about how everything worked.
I remember when they deprecated the library and instead made it a learning resource on implementing auth from scratch. Brilliant decision, much respect to the author.
You could almost call the comparison between Clerk and Better Auth unfair. One is a service and one is a library, apples to oranges. Any third-party service integrated into a stack is a liability, libraries as well, but to a lesser degree. It’s about time for more services to be replaced by libraries. Better Auth really shows how to do that imo, it’s a library that integrates on the frontend, backend, and database. This is why it’s so good.
Using Clerk, quite unhappy with it. No proper RBAC (roles are tied to organizations, not stored on user itself, so you cannot have a concept of global admin or something like that, unless you use metadata for storing arbitrary key value paris), and more than once in the past weeks/months it had a downtime causing the whole app to fail.
Disclaimer: I work for a Clerk competitor, FusionAuth.
Can you share your evaluation process? I'm always curious how folks evaluate auth providers.
Did you do a spike? Full POC across a couple of solutions? Rely on a recommendation from a friend? Run through a quickstart and decide it worked and you had bigger problems to solve? Something else?
i'm currently in the process of evaluating switching our b2b app off clerk too. a customer asked for custom roles (diverging from our default set), which is technically possible on clerk now with role sets, but it's a bit of a workaround rather than first class.
thankfully i'm familiar with better auth from a side project, but migrating SSO/SCIM sounds like it might be a bit of a pain
Had a similar journey recently. Started with Stack Auth, found it unusable in production due to extremely hard rate limits and bad performance even when not rate limited. Switched to WorkOS AuthKit, which works much better and supports useful enterprise features. But inclined to BetterAuth for new projects.
- Syncing external auth provider state with your user state is a bug center. It helps to keep as little state as possible in the auth provider, but there is still some.
- Refreshing JWT access tokens every few minutes is another bug center and honestly there is no need to do this if you control your own auth.
- WorkOS does not have a complete API. It is built on the assumption that you have one product per billing account and a fixed number of environments (staging, production, and they can give you another one if you ask support). You have to whitelist redirect and other URLs in the dashboard, and there doesn't seem to be an easy way for agents to do it.
Outsourcing auth does not make much sense IMO. The less you can split your state over multiple services the fewer problems you will have. Sometimes it is inevitable, like for payments, or if you need specialized databases for performance reasons. But for auth there is really no good reason if good libraries are available. To people who say that using a service will help you get started faster, none of the problems I hit with auth services had to do with having high scale -- most of them hit before I even launched.
> Outsourcing auth does not make much sense IMO. The less you can split your state over multiple services the fewer problems you will have.
I agree with the general principle. Fewer moving pieces make for more stable applications ("choose boring technology"[0]).
However, I was wondering what you do when you have more than one application that the same userbase wants to access. I can see 3 options:
1. make them register/have credentials for each application (not a great user experience)
2. use a standalone auth server and deal with the increased complexity
3. pick one of your applications to 'own auth' and have the other applications delegate to it. congrats, you've just invented a standalone auth server that is coupled to one of your apps
My biggest question, that I didn’t see answered, is how the transition comes to an end. If creds aren’t in your system, how well do they transfer? Does it require user involvement? What if they don’t fast enough? What about complex integrations like SSO, SCIM, and passkeys (which are domain scoped)?
Better auth is great! I love how it's way more hackable than using a something like Clerk. We were able to add a plugin to allow auth via iframe postMessage (embedded in a CRM) and everything worked seamlessly.
Does Better Auth still have the weird design to be everything “request header based”? I remember running admin scripts and tests to be very hacky due to it cause if you skipped that plugins wouldn’t run
If anything I feel like Clerk adoption is becoming the norm in recent years. I started using it about a year ago and found it to have troublesome reliability.
Lol wut? you get all of your auth data in your own db in 1 cli command. You are not tied to any on db provider. On top of that you get hundreds of auth features like oauth providers (I use it to allow users to log in via google, apple, github) and the best part it's free. Not saying Supabase and Clerk are bad, but they cost money. With better auth you pay exactly $0 for all of this.
I’ve looked at these auth providers many times over the years and I just don’t get the value. It takes me a couple of minutes to set up auth. Why would I want a dependency? It doesn’t help me with the hardest part which is configuring Google and Apple sign in stuff on Google and Apple. I just don’t get it.
I use better auth for a side project i'm working on. It's open source, you can pay them to manage your user/auth tables if you want or you can run it all on your own db.
Sure, I guess if I'm using a web framework that is not quite batteries included, that makes sense, but Django, ASP .NET Core, Ruby on Rails, and so many others are open source and have authentication / users / roles baked in out of the box.
I remember Laravel with Socialite [0]. Laravel is what I usually reach for Web SaaS MVP. You only need a VPS and a managed database for testing out the market and can scale a lot without increasing expenses that much..
Lol, I developed for entrepreneurs who mostly wanted a working proof of concept of their ideas. I guess now you can vibecode them with SaaS for core technical needs.
I’m setting up a new system (Auth for user facing app, client facing dashboards, internal tool usage, etc.)… it has anonymous app users who can be upgraded to real users upon Auth, organizations, multi-tenant capabilities, all kinda of sign in options… haven’t written a single line of special code this is all handled by included plugins and defaults. It just works and I own all the data.
> Some important context is that Clerk is a major success. They just raised 50 million dollars and they have lots of satisfied users.
And even more users who are looking to escape. Clerk is just a mess. They are trying to cram EVERYTHING into their libraries: Web3 crap, Stripe, etc. Clerk's JS blob is now triggering the browser inspectors for being slow to load.
Every time when we upgraded React, Clerk libraries were the biggest pain with their transitive dependencies. We had issues with Stripe libraries with conflicting versions, etc.
And forget about debugging it. The libraries are obfuscated, and the TS code is impenetrable mess of abstractions to support "isomorphic" code that can run transparently on the frontend and backend.
And their platform itself is lacking important functionality, like freaking audit logs and versioning. Somebody (probably) accidentally changed a setting in their console, and we couldn't trace back when it happened or who did it.
Edit: oh yeah, and don't forget their unreliability. I had to wake up on Sunday to deal with Clerk failing the API calls for token refreshes last week.
Uhm, companies like Replit and several other large startups are actually adopting Clerk. I guess if your world mainly revolves around X (formerly Twitter), it can seem like everyone is moving away from Clerk.
Also, Better Auth’s X presence is pretty much centered around criticizing every auth provider out there, so the discourse there tends to skew heavily negative.
> Also, Better Auth’s X presence is pretty much centered around criticizing every auth provider out there, so the discourse there tends to skew heavily negative.
This from an account created 2 hours ago, with a username that’s a negation of the BetterAuth founder…
If you’re Clerk stakeholder why not just come out as yourself and engage openly!
Clerk looks _really_ good initially. It's perfect if you want to prototype something and not care about auth.
It's only when you start getting into the details that you begin to suffer. For example, there's _still_ no way to do offline auth on mobile. So that your application could be opened if there's no connectivity at the moment. But hey, you can do the Metamask Web3 blockchain thingie!
I have never used Twitter/X, and I don't even have an account there. I'm purely talking about my personal experience and the experience of other companies that I know personally.
> Also, Better Auth’s X presence is pretty much centered around criticizing every auth provider out there, so the discourse there tends to skew heavily negative.
They are actually not wrong. Auth is not such a hard task, it's just a lot of drudgery that detracts you from the actual goal of your company. But it's critical functionality that MUST ALWAYS WORK, before all else. And Clerk just fails this test.
I'm switching my company to Logto (it's lightweight and when something breaks, I know how to pick up the pieces), so I don't even have an opinion on Better Auth.
We ended up caching the credentials and the JWT refresh token from Clerk, and then manually requesting the access token using the Clerk's sparsely documented frontend API. Except that to do this with Clerk, we needed to fake the cache API and then pluck the token out of the undocumented "__clerk_client_jwt" key.
This is supported by Better Auth out-of-the box. It doesn't hide these kinds of stuff from you.
Is that the new library that came out in rails 8? Saw someone present on that at RailsConf 2025 and it seemed like a great solution for all rails apps. Hope it leads that ecosystem to get rid of devise (which I always found confusing).
Has anyone used Keycloak for actual production? I have often thought about it but I stick to Auth0 just because I don't know if Keycloak has a good track record?
I've seen it used in production by larger orgs. The scale where you plan for around 6 months of migration, customization and integration of your legacy zoo with 7 different user account DBs. On one hand, all of these projects were successful and now run it in production. On the other, they all really needed the 6 months to whip it into shape.
You might be interested in some of the presentations at KeyConf[0]. You can also get some real world stories from the Reddit[1].
I was at KubeCon EU this year (representing my employer, FusionAuth) and there were lots of folks who were running Keycloak who came and chatted with us.
It's a different set of tradeoffs than Auth0 or other SaaS services. More control, but more responsibility too.
Yes! I used keycloak for multitenant auth and it worked fine - a little dated but functional. Nowadays I'd probably stick to something like Clerk/BetterAuth/Supertokens.
Hey, I work at Val Town, and we definitely struggle to describe what it is because the platform is so broad, but I'll try: a javascript code editor in the browser that auto-deploys that code to our servers whenever you edit a file. You can run code by clicking a button, on a cron, via HTTP. And there's other stuff like SQLite and blob storage
So yes, like Cloudflare Workers in some ways. Or like CodePen but fullstack. Or Replit. The val town "founding poem" was:
This comment thread was a nudge to push out a quick experimental change to the val.town landing page, so thank you. Very open to feedback on that messaging if anyone has any. Most customers these days ask about using val town with claude code, so you could also think of us as a deployment platform for vibe coded apps
That's a good question, I was having a hard time figuring that out myself. They call themselves the "zapier" for developers. In reality it seems kinda like a FaaS but idk. They have a code intelligence product that seems like a FIM autocomplete. Very confusing product suite.
Lastly - here's the law of mother nature:
Software funded by VCs will milk you and will milk you dry. Not now, not tomorrow maybe decade or two later when the hot potato being passed around reaches to the last party holding the bag that now needs its money back with all that others had their 10x returns already taken from them by soldng them at an inflated price to the last party holding the bag.
Clerk's pricing should be evidence enough.
SSO, SAML, SCIM, OIDC, OAuth, 2FA, passwordless auth, verification tokens, etc etc, And, variations of each for wildly popular systems you’ll be expected to integrate with but don’t support the exact spec.
For a while at my company, half our support engineers time went to handling random SSO issues that came up in our home built auth system.
All you need is Apple and Google Oauth.
If you are a B2C app, you are probably more concerned about:
- social providers (Apple and Google being the big ones, but others could play a role--FB or Tiktok for example)
- easy registration (but not too easy, you want to avoid bot spam)
- self-service account management (updating profile fields, consents [CCPA, GDPR, others], resetting passwords
- single sign-on between your apps (if you have multiple)
- language support (for your backend, and mobile/web front end)
- cost
- possibly MFA, possibly passkeys
It allowed us to do SSO for small one-off marketing / campaign focused sites. I could give a specific login URL and it would always log you in if you were already logged on.
To expand a bit more: if a business is faced with a choice to save some money by increasing risk, having people who’s job it isn’t managing and supposedly securing that information, or to have a third-party who job is literally to handle and worry about those things, who carries independent insurance, and who is on the hook if they lose customer data, and in exchange the business is simply taking the risk of associating with business that could do a poor job — which of those options sounds more appealing from a business sense? It’s a lot easier to blame someone else than earn back trust for your own major mistakes because you tried to write your own software to save a little money.
That’s the SaaS value proposition.
Okay, so… what are those cases? I’m also curious.
If you're willing to make a third party SaaS's uptime the ceiling for your own org, you can delegate auth. Github might not be a good choice for SSO.
If you're not threatened by per-user-per-month fees, you can delegate auth.
If your threat model is compatible with a third party having visibility into your user's network location and the frequency and duration of their activities across your org, you can delegate auth. (Okta will probably not inform your competitor that your main sales guy is in North Carolina this week and has logged in from the conference room wifi of your competitor's main client.)
If you can trust the third party to not allow an interloper to bypass your requirements, you can delegate auth.
The tl;dr of the article is that there are auth specific features that are not differentiated but that users expect. Just like you might outsource pieces of functionality like data storage and message sending to specialized servers/libraries/applications, you can do the same with authentication.
The article could use some improvements, tbh, it is 2.5 years old.
Whether it's painful to put in later or not is sadly nothing that the managers and executives concern themselves with.
Wrote an article about that here: https://fusionauth.io/articles/authentication/fedcm (hosted at my employer's website)
https://better-auth.com/docs/plugins/jwt
It's worse than that, the combined availability is the product of all components in the critical path. If your software, the authentication layer, and the cloud provider each have 99% availability, and any one of them can bring your service down, then your final availability is just 97%. With eleven components like that you have zero nines of availability.
That's why reducing components and going for reliable solutions is so important. I'm happy that the team took this path.
I remember when they deprecated the library and instead made it a learning resource on implementing auth from scratch. Brilliant decision, much respect to the author.
Would think twice before using it in the future.
Can you share your evaluation process? I'm always curious how folks evaluate auth providers.
Did you do a spike? Full POC across a couple of solutions? Rely on a recommendation from a friend? Run through a quickstart and decide it worked and you had bigger problems to solve? Something else?
thankfully i'm familiar with better auth from a side project, but migrating SSO/SCIM sounds like it might be a bit of a pain
- Syncing external auth provider state with your user state is a bug center. It helps to keep as little state as possible in the auth provider, but there is still some. - Refreshing JWT access tokens every few minutes is another bug center and honestly there is no need to do this if you control your own auth. - WorkOS does not have a complete API. It is built on the assumption that you have one product per billing account and a fixed number of environments (staging, production, and they can give you another one if you ask support). You have to whitelist redirect and other URLs in the dashboard, and there doesn't seem to be an easy way for agents to do it.
Outsourcing auth does not make much sense IMO. The less you can split your state over multiple services the fewer problems you will have. Sometimes it is inevitable, like for payments, or if you need specialized databases for performance reasons. But for auth there is really no good reason if good libraries are available. To people who say that using a service will help you get started faster, none of the problems I hit with auth services had to do with having high scale -- most of them hit before I even launched.
I agree with the general principle. Fewer moving pieces make for more stable applications ("choose boring technology"[0]).
However, I was wondering what you do when you have more than one application that the same userbase wants to access. I can see 3 options:
1. make them register/have credentials for each application (not a great user experience)
2. use a standalone auth server and deal with the increased complexity
3. pick one of your applications to 'own auth' and have the other applications delegate to it. congrats, you've just invented a standalone auth server that is coupled to one of your apps
What am I missing?
0: https://boringtechnology.club/
Anyone remember Auth0 and passportjs?
The churn of auth services is never ending, but I suppose so are the standards.
Heya, Auth0 is still around! They got bought by Okta in 2021 but still have a free tier and we see them in a lot of bake-offs.
https://laravel.com/docs/13.x/authentication
https://laravel.com/docs/13.x/socialite
https://laravel.com/docs/13.x/sanctum
[0]: https://laravel.com/docs/13.x/socialite
And even more users who are looking to escape. Clerk is just a mess. They are trying to cram EVERYTHING into their libraries: Web3 crap, Stripe, etc. Clerk's JS blob is now triggering the browser inspectors for being slow to load.
Every time when we upgraded React, Clerk libraries were the biggest pain with their transitive dependencies. We had issues with Stripe libraries with conflicting versions, etc.
And forget about debugging it. The libraries are obfuscated, and the TS code is impenetrable mess of abstractions to support "isomorphic" code that can run transparently on the frontend and backend.
And their platform itself is lacking important functionality, like freaking audit logs and versioning. Somebody (probably) accidentally changed a setting in their console, and we couldn't trace back when it happened or who did it.
Edit: oh yeah, and don't forget their unreliability. I had to wake up on Sunday to deal with Clerk failing the API calls for token refreshes last week.
Uhm, companies like Replit and several other large startups are actually adopting Clerk. I guess if your world mainly revolves around X (formerly Twitter), it can seem like everyone is moving away from Clerk.
Also, Better Auth’s X presence is pretty much centered around criticizing every auth provider out there, so the discourse there tends to skew heavily negative.
This from an account created 2 hours ago, with a username that’s a negation of the BetterAuth founder…
If you’re Clerk stakeholder why not just come out as yourself and engage openly!
It's only when you start getting into the details that you begin to suffer. For example, there's _still_ no way to do offline auth on mobile. So that your application could be opened if there's no connectivity at the moment. But hey, you can do the Metamask Web3 blockchain thingie!
I have never used Twitter/X, and I don't even have an account there. I'm purely talking about my personal experience and the experience of other companies that I know personally.
> Also, Better Auth’s X presence is pretty much centered around criticizing every auth provider out there, so the discourse there tends to skew heavily negative.
They are actually not wrong. Auth is not such a hard task, it's just a lot of drudgery that detracts you from the actual goal of your company. But it's critical functionality that MUST ALWAYS WORK, before all else. And Clerk just fails this test.
I'm switching my company to Logto (it's lightweight and when something breaks, I know how to pick up the pieces), so I don't even have an opinion on Better Auth.
Does Better Auth offer this? Or any other auth libraries or solutions? I haven't heard of any, but haven't done an intensive look either.
I suppose you could do something with a cached JWT or cached password hash (though sending a password hash to a mobile client spooks me).
I'm in the space and interested in learning more.
This is supported by Better Auth out-of-the box. It doesn't hide these kinds of stuff from you.
- https://github.com/agoodway/introspex (generate Ecto Schemas from postgres tables)
- https://github.com/agoodway/pgrest (Supabase/PostgREST compatible query engine)
I also found this helpful in the migration: https://github.com/supabase-community/supabase-ex
Nothing for auth, I basically did a one-off script for that. Phoenix auth stuff that comes out of the box is great.
no need for 3rd party provider.
I was at KubeCon EU this year (representing my employer, FusionAuth) and there were lots of folks who were running Keycloak who came and chatted with us.
It's a different set of tradeoffs than Auth0 or other SaaS services. More control, but more responsibility too.
0: https://events.linuxfoundation.org/kubecon-cloudnativecon-eu...
1: https://www.reddit.com/r/KeyCloak/
Easy to use and high reliability. Some of these other providers are not the best at reliability.
Maybe you can help me out: I still have no idea what val.town is. I guess it's an alternative to Cloudflare Workers?
So yes, like Cloudflare Workers in some ways. Or like CodePen but fullstack. Or Replit. The val town "founding poem" was:
> If GitHub Gists could run,
> And AWS Lambda were fun