codingismy11to7

@MichaelArnaldi @YuvarajMadineni sorry you thought I wasn't being constructive. you're right, just porting the code would be easier

That's not really the way to interact positively, a few things: 1. Effect v4 is in BETA and we are looking for feedback, as explained it is not certain if Effect.Service will be ported or not. 2. Effect v3 is maintained for at least a year. 3. You built on something that was flagged as experimental #L13577" target="_blank" rel="nofollow noopener">github.com/Effect-TS/effe… 4. There are ways to migrate via Service.use without the need for accessors. 5. Worst case you can port the Effect.Service code and it will work just fine. That said if you want to remove Effect that's fine, gonna be way more than a 2k lines PR, updating major versions is by definition breaking, we have nothing to do with your choice of Scala decades ago.

Before GTA 6!

@MichaelArnaldi @YuvarajMadineni I wasn't trying to be negative at all, I was panicked so I'm sorry. yes we can port all the code over ourselves. that's probably what we'd do instead of dropping effect. but still I'd rather not see Service die

@GiulioCanti lol I mean, I'm super happy with effect schema 3 😉

I am so happy with Effect Schema v4 that I don't think I will ever write another schema library. Maybe 😄 x.com/EffectTS_/stat…

@MichaelArnaldi @YuvarajMadineni for us that means dropping effect I think. which will be real hard after the next two refactors (schema & http) go in. so basically actually we have to decide now. I've already been told once this month by the CEO that I made a terrible choice in choosing Scala 15 years ago

@MichaelArnaldi @YuvarajMadineni I mean... our server cleanly uses hundreds of Services. accessors are used heavily in test code. It was already a pretty big deal to get us to adopt effect, if I'm coming in and saying "sorry here's another 2k file MR because I built on a seemingly solid surface that wasn't..."

@unclebobmartin Agree. First you get a thrill, but then you realize your job is now conversing with a super powerful toddler genius who needs babysitting and constant repetition. You get tired of “talking”. Not everybody is gonna like that job.

@MichaelArnaldi omg please tell me this is effect-native oauth and I get to delete my home grown implementation? and that it works seamlessly with effect-platform's http builder? 😄

Progress to be followed at: github.com/mikearnaldi/ac…

Differently from what people think for me automated software development enables more correct design decisions and avoids lazy behaviour, for example for Accountability I am now implementing authorization, I would have never implemented full ABAC myself but I am doing it now

@kitlangton we seem to be following similar tech-life arcs - have you also been nix-pilled? anyway here's my nvim setup for effect github.com/codingismy11to…

@kitlangton @dillon_mulroy haha I noticed that, I've also left the warm familiar arms of IDEA for nvim

A little guilty pleasure for the holidays...

@imax153 @EffectTS_ so excited for the day we get to move off my decrepit stack of json schema hacks

Like I said before, you all are NOT READY for @EffectTS_ Schema v4

@gitlab who can I talk to about recovering my gitlab.com account? I "successfully" did the automated recovery and am still denied login with new pw. not sure where to go if I can't sign in.

@kevinleequinn I also added temp to waybar yesterday

@dhh I installed this weekend and it took less than three minutes. added some packages and I'm off to the races. need to get my media keys working, temp in the waybar and replace ChatGPT with Grok...

Please keep posting and keep tagging! It's such a delight to see all these machines running Omarchy 😍. I've poured hundreds and hundreds of hours into this project, so seeing folks appreciate the work is very special. Never apologize for being excited about computers!

@kitlangton @EffectTS_ i guess that's not the point of this, but I got our entire backend ported to effect and I've only had time to do one teaching session and I didn't really cover the layers at all, just told them to read the docs. if I could send some team leads to this....

@kitlangton @EffectTS_ hey so I'm all good on these but like...my team would be fascinated. could I send a representative or two?

Would anyone be interested in pre-alpha testing some @EffectTS_ lessons? This would involve me screen-sharing with, ideally, a few 𝚎𝚏𝚏𝚎𝚌𝚝 novitiates, and attempting to teach them some aspects of the library. First up is likely something on 𝙲𝚘𝚗𝚝𝚎𝚡𝚝 and 𝙻𝚊𝚢𝚎𝚛.

@MichaelArnaldi @WickyNilliams @CampuzanoJoe @kitlangton @SubtleGradient ☝🏻can confirm, just got done with that final cleanup on our codebase. ~550 changed files, 2-300 effect services. still a ways to go, though, until we're able to switch to -schema and then (finally!) -http. but the team is already excited

@WickyNilliams @CampuzanoJoe @kitlangton @SubtleGradient in large codebases you usually start with leafs indeed, sometimes a mixture of top-down and bottom-up but it will always be incremental, usually after a while a lot of code becomes effect and one final effort to clean-up everything can be done

In response to the good @SubtleGradient, who recently asked [paraphrasing] will Effect-TS avoid RxJS’s “flash-in-the-pan” fate in the JS ecosystem? RxJS was an early solution to the challenges of async and reactivity. Gradually, aspects of this problem were solved by simpler, more imperative apis: React, hooks, react-query, async/await, etc. The Reactive Streaming paradigm has failed to really take off in any language—e.g., a few years back, Apple released its very own first class reactive streaming API called Combine, which was deprecated within a year, supplanted by a much simpler, more imperative API. It also doesn’t help that RxJS is inspired by Java’s Reactive Extensions, and thus its API surface area is circumscribed by existing, non-great Rx conventions. For whatever reason, however theoretically elegant, modeling everything as a reactive stream appears to suck (as objectively judged by the market). This inherent suckiness, combined with competition from more modern alternatives, has driven people away from RxJs. Now, with Effect, there isn’t anything in JavaScript that solves these same problems, let alone more simply. If you want resource safety, retries, log spans / annotations, typed errors, composable dependency injection, interruptibility, and all the other benefits derived from Effect’s programs-as-values design, you’ll either have to use Effect or roll your own much shittier ad hoc version, once per codebase. Promise isn’t really an alternative to Effect, as it merely provides barebones async support and none of the aforementioned capabilities. Once you’re reaching for AbortController, you’re doomed. The only true alternative is total acquiescence to a diminished quality of life: you won’t know all the ways your program might fail; parallel computations won’t be interrupted; retrying will be ad hoc and verbose; logging shmlogging; telemetry shmelemetry. The one convincing argument contra-Effect is that you just don’t need or want to solve these problems. In many apps, this might be fine, if you can tolerate a certain amount of pain and turmoil. Clearly, this approach works and can scale, it is indeed the approach of most companies. The question is: can it scale as fast, as far, and with as small of a team of developers? I’d reckon the difference is measured in orders of magnitude. In my experience, once you do get on board with Effect, understand what it offers, and acclimate to its syntactic idiosyncrasies, you could not possibly fathom going back. I imagine it would be like having an 18th-century doctor struck with a prophetic vision about Germ Theory before returning to his 18th-century operating room. He might then feel, justifiably so, that he was doing great harm to the people he was supposed to be helping.

@GiulioCanti @Patrick_Roza @jdegoes wonderful, another tool in the toolbox without having to add another library

@Patrick_Roza @codingismy11to7 @jdegoes To be precise, the `Optic` module works only with plain objects. The gap with any custom type is bridged by the `ToOptic` module through isomorphisms: #generating-an-optic-from-a-schema" target="_blank" rel="nofollow noopener">github.com/Effect-TS/effe…

THEY DID IT.💪🎉

@MichaelArnaldi I know it fucks up on effect code constantly 😆

it's pretty clear to me that the development world is currently divided between the ones who have seen what AI can do and those who didn't, and the conversation between the two groups is almost useless as they don't share the same objective reality

@twiabp @FoolOfKingLear @Bandcamp haha yeah that's where I preordered for the exclusive variant (assume it was exclusive anyway)

@FoolOfKingLear @Bandcamp It’s more of a reminder that we have the vinyl variant up there, and proving that people do in fact order things on there

Recently a friend of mine said no one uses Bandcamp. Said we could have leaked the record on there, and no one would have even noticed. Prove him wrong by ordering the record through @Bandcamp

the world is a gimp-footed dog and I am no longer bark bark bark