Sixstring982

@alessandrod You're my hero

not the best, I had a 55s restart yesterday, but consistently getting <1m now

@theo Wait until you `RUSTFLAGS="-C target-cpu=native" build --release`

Rust build times are so bad holy shit

@theo @t4styycs IMO it depends on whether or not you're going to use `activeTodos` in a dependency array, and need a stable reference. I would probably `useMemo` here just to keep the reference stable, because bugs caused by unstable references are really annoying

@t4styycs Describe in detail what you think adding a “useMemo” here would do, I’m curious

I will never forget the life of me understand why people write code like this

@AndersonAndrue Hey awesome, I find this syntax much more manageable than trying to cram everything into the first argument of `fc.property` -- especially stuff that depends on each other. I guess I shouldn't be surprised, since FC is monadic; but still!

Finally, we write some tests! Effect has an awesome built-in integration with FastCheck for arbitrary data generation from Schemas!

Why @EffectTS_ ? Testing! Effect's built-in capabilities for property-based testing make it simple to generate test data and test edge cases you'd never think about. Use Effect. How? 👇🧵

@dillon_mulroy If you aren't jinja-templating your nginx configs, are you really chilling?

😏

it's an nginx and chill kinda night

@janbhwilhelm @dillon_mulroy My guess is that this is a property test, though the thing being deserialized could also just be really big

@dillon_mulroy just curious, why would tests that e.g. decode Uint8Arrays take 910ms?

dns packet decoding and encoding is done with the exception of name (de)/compression

@dillon_mulroy That is one fat stack lol

if you're not stacking prs, you're doing it wrong

Four months ago we took down our highly-praised website and replaced it with simple monospace text-heavy markdown files. Since then, people have reached out almost weekly asking what impact this had. Did our conversions tank? Have we just completely given up? Let's dig in.

my favorite part of the @arktypeio docs

@dillon_mulroy @EffectTS_ That `concurrency: 25` threw me off a bit -- I suppose `enrichEntityWithMetadata` is something that returns a `Promise` under the hood? Good stuff here, I need to try out EffectTS!

you love to see it (it is @EffectTS_ )

@welltypedwitch Awesome article, I found it today in Haskell Weekly! TIL about Linearity -- I'm excited to see that in Haskell. Its' a pretty cool feature!

@dillon_mulroy @MiTypeScript Hey cool, that seems like a useful trick! I'm gonna need to look for places to use it lol

This is such a good explainer of distributive conditional types in TS and worth your time to watch - kudos @MiTypeScript 🔥 youtube.com/watch?v=DNTipH…

@trikcode Adapter pattern

Explain this in programming terms

@devagrawal09 Indeed! Monads are really useful for encoding DSLs. You could totally make a builder monad, which could be useful in certain situations!

I'm seeing some parallels between the builder pattern in OOP and monads in FP Might be a stretch, but both patterns are designed to give the programmer the ability to create an intermediate structure that represents an operation, compose additional behaviors on top, and delegate the actual operation to a runtime The primary difference being builder objects being mutable, but if every method on a builder just returns a new builder with the new behavior, you basically just have a monad

@sabine_s_ @ocaml_org @techsavvytravvy @typescript Oh cool, OCaml has dynamic arrays now? I guess I never really checked if it did lol But yeah those are usually more performant than linked lists due to better cache locality and whatnot

@Sixstring982 @ocaml_org @techsavvytravvy @typescript if u use a linked list in JavaScript it will be slow 🤷 so why not compare to the new Dynarray?

ok so i'm still implementing parts of @ocaml_org standard library, and obviously the source has a lot of recursive functions which is fine in ocaml bc they're optimized but not so great in js/ts.. should i change the implementations to be imperative, or should i do a system of trampolines and thunks? ...or should i yolo recursive calls and hope js engines figure it out

@ocaml_org @techsavvytravvy @typescript Lol well OCaml would use a linked list and take advantage of tail call optimization, so the recursive function here is a great way to do it!

@Sixstring982 @techsavvytravvy @typescript why not a for loop in OCaml?

@t3dotgg Certainly

...this is definitely AI-generated, right?

@VicVijayakumar @DanielW_Kiwi @Shreyassanthu77 +1, a mat is critical for a standing desk. I've been using a standing desk for nearly 8 years now, and I haven't owned a desk chair during that time. You need a good mat, otherwise it doesn't work.

@DanielW_Kiwi @Shreyassanthu77 amzn.to/4b6Vl80

It's taken a week or two but now the standing desk is starting to get uncomfortable.