I don't know, as far as I can tell the demo on the homepage is bugged? Or maybe I misunderstand how it's supposed to work? If I click reset after 2 seconds and then click add 1 right after it, the reset never fires. What's up with that, that's not what I expected to happen?
Yeah, I can see how some scenarios might want that behavior, but it sort of defeats the purpose of the demo. It wouldn't hurt to have the reset be cancelable too if you wanted to really showcase how it should work
I guess it comes from the `tagsExhaustive` keyword. But I tried to search what it does in the documentation, and it doesn't return any hit. I then searched in the git repository and found it in some examples and CLAUDE.md but it doesn't really seem straightforward.
Perhaps it makes sense for AI agents, but as a human, I will pass.
The project seems super interesting and I've been following since the creator went on an effect podcast I listened to. My only gripe/concern is the fact that the docs are so glaringly AI (likely claude) generated which is something I've come to not expect from technical documentation writing specifically. All in all though, the verbosity tradeoff for explicit design choice here is the right one and I think it does better than elm by remaining in javascript land rather than building a nicer abstraction layer on top that often needs to reach in to do interesting things with the web apis.
Just wondering how a matching backend framework for correctness would look like. Ideally something that doesn't repeat the React mess with two routers but fits right in Inertia style.
It took me 3 months to go from “I will never use it” to “I will never not use it”. Plus it’s always fun to watch the creators of Effect get confused about which brackets to where. The syntax is archaic, luckily we don’t write syntax by hand any more.
I use it from years, it's the only sane way to write typescript.
The learning curve is steep tho.
In any case it's a hard technology to sell, you don't appreciate it from hello worlds. In fact you hate it for trivial programs.
It shows it's benefits at scale.
Lots of tools like T3 and opencode use it at their core because they solve non trivial problems made of queues, retries, dependency injection, schemas, etc.
I love both effect and elm, but I cannot swallow yet another Vue/React crap even if it adds effect niceties.
Especially after having used ruby and elixir extensively after years of react/Vue, it feels so backwards (and LLM unfriendly) to split front and backend unless you have gargantuan reactivity needs (you don't).
I wish JS offered just one proper server side focused frontend solution.
Is it just me? But every time I use a paradigm of global immutable state, almost always I run into edge cases where practically it falls apart either semantically (mental model explosion) or creates performance issues, whether they stem from architectural problems with the framework itself, or just the way computers operate.
As agentic coding matures, this kind of project is the right direction. The mechanics of writing the code become less important. But having a language framework that naturally resists mistakes will become increasingly useful and important.
Similar to how Rust is obviously a better choice today than C++. Who cares if the language is “more difficult” to code in, if the agents are doing the coding. What we need is building blocks that make it harder to author bugs.
That's an interesting - and maybe correct - point. Looking at the examples, it seems very long winded and not something I would want to write myself - but maybe I don't need to care about that anymore, as you say!
This time around in the pre-covid days, F# had a surge of popularity, there was Fable and Giraffe and an elm-style set of frontends (the generic pattern is called MVU - Model View Update), it didn't take off because A) we are still a trend/fad-driven industry and B) there was no AI to help hold your hand in refactoring all of your apps.
These days? Doing this FE SPA framework for Effect/typescript familiarity or Fable for F# familiarity and a model-view update pattern can help one do CQRS better, force you to think about stronger Hexagonal Architecture patterns, and overall help your application care about long term correctness more without you having to hand code every discriminated union as such.
The LLM can't replace you, but it can be an enzyme that lowers the activation energy required to switch from bad old MVC mutable tag soup to MVU frontends with strong data interface designs held in sqlite-dbs-per-aggregate (if you are of a DDD or event sourcing taste, which this pattern does not require!)
I was thinking about agentic coding as well and a few years ago I probably would have really been interested in this project. But now it sits at a lower abstraction where I care less since the agent is the one managing it. It's strange
I don't know, as far as I can tell the demo on the homepage is bugged? Or maybe I misunderstand how it's supposed to work? If I click reset after 2 seconds and then click add 1 right after it, the reset never fires. What's up with that, that's not what I expected to happen?
It's a framework for correctness, after all :)
Yep, it seems that if you hammer messages things get easily dropped or the async behavior is not executed.
IMO the reset should fire regardless what happens after it.
Agreed - I hit "Reset after 2 seconds", I hit "Add 1", and sure - it adds one, but it also unexpectedly cancels the reset.
Which I don't expect from the description, and I also don't expect from the code.
---
So outside of the simple "This is broken..." feedback, I want to further pick on this example:
Don't fucking hide the imports.
Especially don't fucking hide the imports if you're showing example code, and you're doing things like
```
import { Match as M, Schema as S } from 'effect'
import { m } from 'foldkit/message'
```
Which I only know because I bothered to dig through the example playground counter (which is a different example entirely!)
It's a huge issue to show demo code where concepts magically appear, and it's just bad manners to use shorthand at the same time.
Yeah, I can see how some scenarios might want that behavior, but it sort of defeats the purpose of the demo. It wouldn't hurt to have the reset be cancelable too if you wanted to really showcase how it should work
I guess it comes from the `tagsExhaustive` keyword. But I tried to search what it does in the documentation, and it doesn't return any hit. I then searched in the git repository and found it in some examples and CLAUDE.md but it doesn't really seem straightforward.
Perhaps it makes sense for AI agents, but as a human, I will pass.
The whole setup reminded me of the mess people would make with RxJS, and this comment chain seems to indicate it's got the same drawbacks.
The project seems super interesting and I've been following since the creator went on an effect podcast I listened to. My only gripe/concern is the fact that the docs are so glaringly AI (likely claude) generated which is something I've come to not expect from technical documentation writing specifically. All in all though, the verbosity tradeoff for explicit design choice here is the right one and I think it does better than elm by remaining in javascript land rather than building a nicer abstraction layer on top that often needs to reach in to do interesting things with the web apis.
This feels a lot like sagas and redux sagas. Anyone else got that vibe?
I was feeling similarly: epics with redux-observable but swapping RxJS for Effect.
Its the same "solved" pattern, but with the pattern being more deeply rooted in a project.
Just wondering how a matching backend framework for correctness would look like. Ideally something that doesn't repeat the React mess with two routers but fits right in Inertia style.
That is a fair question. Probably one could inspiration by https://github.com/lustre-labs/lustre
How is Effect, for those who have used it? It's good that algebraic effects are becoming more mainstream.
It took me 3 months to go from “I will never use it” to “I will never not use it”. Plus it’s always fun to watch the creators of Effect get confused about which brackets to where. The syntax is archaic, luckily we don’t write syntax by hand any more.
It’s great and the deep compiler integration means with a skill the agents write it well
I use it from years, it's the only sane way to write typescript.
The learning curve is steep tho.
In any case it's a hard technology to sell, you don't appreciate it from hello worlds. In fact you hate it for trivial programs.
It shows it's benefits at scale.
Lots of tools like T3 and opencode use it at their core because they solve non trivial problems made of queues, retries, dependency injection, schemas, etc.
I've written a vscode extension with it:
https://github.com/dearhuman/effect-decorate/blob/main/src/e...
T3? Isn’t T3 just a TRPC boilerplate kinda thing?
If only I could have the syntactic beauty and brevity of Svelte 4 with the correctness of this <3
This looks interesting, and I've been keeping an eye on this for a while - but does anyone have experience from actually using this for real?
I love both effect and elm, but I cannot swallow yet another Vue/React crap even if it adds effect niceties.
Especially after having used ruby and elixir extensively after years of react/Vue, it feels so backwards (and LLM unfriendly) to split front and backend unless you have gargantuan reactivity needs (you don't).
I wish JS offered just one proper server side focused frontend solution.
Is it just me? But every time I use a paradigm of global immutable state, almost always I run into edge cases where practically it falls apart either semantically (mental model explosion) or creates performance issues, whether they stem from architectural problems with the framework itself, or just the way computers operate.
its a trade off for sure
but my experience working with elm is that things just work almost always the first time
if they can bring that over, many will take that trade off
yes somethings are harder to express, but its subjective if its a bad thing given the correctness guarantees
i haven't firmly run into the perf ceiling myself, but yes it is obviously there for more interactive apps
Finally someone said it
As agentic coding matures, this kind of project is the right direction. The mechanics of writing the code become less important. But having a language framework that naturally resists mistakes will become increasingly useful and important.
Similar to how Rust is obviously a better choice today than C++. Who cares if the language is “more difficult” to code in, if the agents are doing the coding. What we need is building blocks that make it harder to author bugs.
That's an interesting - and maybe correct - point. Looking at the examples, it seems very long winded and not something I would want to write myself - but maybe I don't need to care about that anymore, as you say!
This time around in the pre-covid days, F# had a surge of popularity, there was Fable and Giraffe and an elm-style set of frontends (the generic pattern is called MVU - Model View Update), it didn't take off because A) we are still a trend/fad-driven industry and B) there was no AI to help hold your hand in refactoring all of your apps.
These days? Doing this FE SPA framework for Effect/typescript familiarity or Fable for F# familiarity and a model-view update pattern can help one do CQRS better, force you to think about stronger Hexagonal Architecture patterns, and overall help your application care about long term correctness more without you having to hand code every discriminated union as such.
The LLM can't replace you, but it can be an enzyme that lowers the activation energy required to switch from bad old MVC mutable tag soup to MVU frontends with strong data interface designs held in sqlite-dbs-per-aggregate (if you are of a DDD or event sourcing taste, which this pattern does not require!)
I was thinking about agentic coding as well and a few years ago I probably would have really been interested in this project. But now it sits at a lower abstraction where I care less since the agent is the one managing it. It's strange
And tla+ / sequence diagram / user flow diagrams / data models all in one "thing"