Alice ✨
@welltypedwit.ch
There’s no sense crying over every mistake. You just keep on trying ’til you run out of cake
- GLaDOS
https://welltypedwit.ch
- GLaDOS
https://welltypedwit.ch
Reposted by Alice ✨
Next Rust in Paris speaker is @wffl.ihatereality.space, member of the Rust compiler team.
Their talk will be: "Little rust delights" where they will talk about some nice but lesser known tricks and tips in Rust.
www.rustinparis.com
To buy your ticket: ti.to/xperhub/rust...
See you there!
Their talk will be: "Little rust delights" where they will talk about some nice but lesser known tricks and tips in Rust.
www.rustinparis.com
To buy your ticket: ti.to/xperhub/rust...
See you there!
Rust In Paris 2026
Rust in Paris is a community conference for developers passionate about building secure, concurrent, high-performing software with Rust.
www.rustinparis.com
February 10, 2026 at 1:18 PM
Next Rust in Paris speaker is @wffl.ihatereality.space, member of the Rust compiler team.
Their talk will be: "Little rust delights" where they will talk about some nice but lesser known tricks and tips in Rust.
www.rustinparis.com
To buy your ticket: ti.to/xperhub/rust...
See you there!
Their talk will be: "Little rust delights" where they will talk about some nice but lesser known tricks and tips in Rust.
www.rustinparis.com
To buy your ticket: ti.to/xperhub/rust...
See you there!
wait why? you get different tokens that way don't you?
February 9, 2026 at 2:44 PM
wait why? you get different tokens that way don't you?
i love the distance between "american progressive democrats" and "progressive american democrats"
very people's front of america
very people's front of america
February 9, 2026 at 2:37 PM
i love the distance between "american progressive democrats" and "progressive american democrats"
very people's front of america
very people's front of america
intercalate :: forall a. as -> ass -> as
uh oh
uh oh
February 9, 2026 at 2:32 PM
intercalate :: forall a. as -> ass -> as
uh oh
uh oh
(and ocamllsp continuing and producing garbage types after non-type errors doesn't exactly help)
February 9, 2026 at 2:29 PM
(and ocamllsp continuing and producing garbage types after non-type errors doesn't exactly help)
no, it makes sense if you read code outside of your editor or if you don't write type-correct code on the first try and have to deal with the mess of error messages and confusing half-broken inferred signatures
February 9, 2026 at 2:29 PM
no, it makes sense if you read code outside of your editor or if you don't write type-correct code on the first try and have to deal with the mess of error messages and confusing half-broken inferred signatures
(and that just makes sense. rust generates really terrible llvm where every single variable is assigned a stack slot and then relies on LLVM's mem2reg to turn that into something sensible (which is the intended way to do it) but that doesn't happen in debug mode)
godbolt.org/z/MxjaGs8GG
godbolt.org/z/MxjaGs8GG
Compiler Explorer - Rust (rustc 1.93.0)
// Type your code here, or load an example.
// As of Rust 1.75, small functions are automatically
// marked as `#[inline]` so they will not show up in
// the output when compiling with optimisations....
godbolt.org
February 8, 2026 at 9:58 PM
(and that just makes sense. rust generates really terrible llvm where every single variable is assigned a stack slot and then relies on LLVM's mem2reg to turn that into something sensible (which is the intended way to do it) but that doesn't happen in debug mode)
godbolt.org/z/MxjaGs8GG
godbolt.org/z/MxjaGs8GG
debug mode is extremely slow. i've had it make the difference between something taking <100ms and ~10s before
February 8, 2026 at 9:55 PM
debug mode is extremely slow. i've had it make the difference between something taking <100ms and ~10s before
this was surprisingly good (it's about ai for programming in case you don't want to read that)
www.lelanthran.com/chap14/conte...
www.lelanthran.com/chap14/conte...
C and Undefined Behavior
www.lelanthran.com
February 8, 2026 at 8:17 PM
this was surprisingly good (it's about ai for programming in case you don't want to read that)
www.lelanthran.com/chap14/conte...
www.lelanthran.com/chap14/conte...
"issue tracker"? oh you mean RLHF?
February 8, 2026 at 1:07 PM
"issue tracker"? oh you mean RLHF?
Reposted by Alice ✨
Ragebaiting anti-ai compiler devs by calling the lexer the tokenizer, parser decoder and code gen inference.
February 8, 2026 at 3:55 AM
Ragebaiting anti-ai compiler devs by calling the lexer the tokenizer, parser decoder and code gen inference.
it doesn't if you do `echo "$(ps)" | grep xyz`!
February 8, 2026 at 2:27 AM
it doesn't if you do `echo "$(ps)" | grep xyz`!
okay but a ton of rust impls fail the paterson conditions.
like, none of these instances would be accepted (because they all either mention an associated type in their entailed constraints or a type variable (A) that doesn't appear in the head)
like, none of these instances would be accepted (because they all either mention an associated type in their entailed constraints or a type variable (A) that doesn't appear in the head)
February 7, 2026 at 1:50 PM
okay but a ton of rust impls fail the paterson conditions.
like, none of these instances would be accepted (because they all either mention an associated type in their entailed constraints or a type variable (A) that doesn't appear in the head)
like, none of these instances would be accepted (because they all either mention an associated type in their entailed constraints or a type variable (A) that doesn't appear in the head)
oh interesting! but then (-) isn't the additive inverse anymore is it? how do you deal with that mathematically?
February 7, 2026 at 12:34 PM
oh interesting! but then (-) isn't the additive inverse anymore is it? how do you deal with that mathematically?
have you talked to lisp or lean people? ^^
February 7, 2026 at 12:30 PM
have you talked to lisp or lean people? ^^
should it though? like, the paterson conditions are sometimes useful for catching small bugs where you accidentally made something recursive that shouldn't have been but for anything non-trivial you'll usually turn them off with Undecidableinstances anyway
February 7, 2026 at 12:27 PM
should it though? like, the paterson conditions are sometimes useful for catching small bugs where you accidentally made something recursive that shouldn't have been but for anything non-trivial you'll usually turn them off with Undecidableinstances anyway
i mean sure but package management doesn't have that much to do with the module system does it?
February 7, 2026 at 1:35 AM
i mean sure but package management doesn't have that much to do with the module system does it?
hot take: extensible syntax is bad and macros are usually worse than more limited forms of metaprogramming like GHC.Generics/compile time reflection (whatever that actually means) or simpler splices
February 7, 2026 at 1:25 AM
hot take: extensible syntax is bad and macros are usually worse than more limited forms of metaprogramming like GHC.Generics/compile time reflection (whatever that actually means) or simpler splices
you could have a really fancy object as the additive identity that (otherwise) doesn't behave like 0 in R at all, but you would still use the symbol 0 to refer to it in this context
February 7, 2026 at 1:12 AM
you could have a really fancy object as the additive identity that (otherwise) doesn't behave like 0 in R at all, but you would still use the symbol 0 to refer to it in this context
no. in detail: by congruence of equality and the definition of - as addition with the additive inverse, those two statements imply 0 = a - a = a + (-a). by definiton, adding something to its additive inverse returns the additive identity and the additive identity is spelled 0.
February 7, 2026 at 1:12 AM
no. in detail: by congruence of equality and the definition of - as addition with the additive inverse, those two statements imply 0 = a - a = a + (-a). by definiton, adding something to its additive inverse returns the additive identity and the additive identity is spelled 0.
the tweening snapshot test is so cool!
i should really add some interactivity to vega's test suite... ^^ (i'm effectively using snapshot testing except that i'm manually copying the error messages to the right file currently)
i should really add some interactivity to vega's test suite... ^^ (i'm effectively using snapshot testing except that i'm manually copying the error messages to the right file currently)
February 6, 2026 at 5:36 PM
the tweening snapshot test is so cool!
i should really add some interactivity to vega's test suite... ^^ (i'm effectively using snapshot testing except that i'm manually copying the error messages to the right file currently)
i should really add some interactivity to vega's test suite... ^^ (i'm effectively using snapshot testing except that i'm manually copying the error messages to the right file currently)
(oops that got cut off)
it's just that this doesn't really happen for maps because the path of least resistance is using horrible "data structures" like association lists instead
it's just that this doesn't really happen for maps because the path of least resistance is using horrible "data structures" like association lists instead
February 6, 2026 at 9:58 AM
(oops that got cut off)
it's just that this doesn't really happen for maps because the path of least resistance is using horrible "data structures" like association lists instead
it's just that this doesn't really happen for maps because the path of least resistance is using horrible "data structures" like association lists instead
but surely modular implicits will Fix Everything when they release in 2096
February 6, 2026 at 9:53 AM
but surely modular implicits will Fix Everything when they release in 2096
eg. `module T(M : functor(X : S) -> S) = M(A)` probably doesn't behave how you expect
and somehow the module errors are even worse than ocaml's regular type errors
and somehow the module errors are even worse than ocaml's regular type errors
February 6, 2026 at 9:53 AM
eg. `module T(M : functor(X : S) -> S) = M(A)` probably doesn't behave how you expect
and somehow the module errors are even worse than ocaml's regular type errors
and somehow the module errors are even worse than ocaml's regular type errors