home.social

#nusa — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #nusa, aggregated by home.social.

fetched live
  1. Ahhh, #Tuchel brings #James to disrupt the circles of #Nusa.
    Will this succeed? I don't think...

    #NORENG

  2. #Calciomercato | Antonio #Nusa era stato trattato dal #Napoli lo scorso anno. #Manna l’ha sempre tenuto d’occhio e il nome potrebbe tornare di moda @[email protected]

    gianlucadimarzio.com/napoli-nusa-no…

  3. 🚨Now official: Antonio #Nusa, new winger of RB Leipzig!

    … with a contract signed until 2029 ✔️🇳🇴

  4. 🚨🔴 DONE DEAL: Antonio #Nusa will join RB Leipzig with immediate effect ✔️

    ▫️Medical and signing today
    ▫️Transfer fee: around €20-22.5m
    ▫️Contract until 2029.

    19 y/o winger from Brügge will replace Dani #Olmo. All the other RB options as Bakayoko, Conceicao and Cherki are off. Partially too expensive and no agreements have been reached despite new talks.

    @[email protected] | @[email protected] | @[email protected] 🇳🇴

  5. 🚨🔴 Antonio #Nusa, ok the verge to join @[email protected] with immediate effect to replace Dani Olmo ✔️

    19 y/o winger from FC Brügge already on-site to undergo his medical today as per @[email protected].

    Deal will happen.

    @[email protected] | @[email protected] 🇳🇴

  6. devlog

    I mostly complain that all system programming languages locked to C callconv & ABI which make it hard for me to target wasm32-web.

    Now here another deep dive why C ABI cause problems:
    - faultlore.com/blah/rust-layout
    - mcyoung.xyz/2024/04/17/calling

  7. #nusa devlog

    I mostly complain that all system programming languages locked to C callconv & ABI which make it hard for me to target wasm32-web.

    Now here another deep dive why C ABI cause problems:
    - faultlore.com/blah/rust-layout
    - mcyoung.xyz/2024/04/17/calling

    #wasm #webassembly #c #clang #rustlang

  8. devlog

    alright, so `__attribute__((address_space(N))` has something todo with wasm multi-memory 🤔

    openasip.org/user_manual/TCE/n

  9. devlog

    I wonder if there's `rustc` flag to bypass proper ABI adjustment 🤔

    I want to target wasm32 but the C ABI Convention get in the way.


    github.com/rust-lang/rust/issu

  10. #nusa devlog

    I wonder if there's `rustc` flag to bypass proper ABI adjustment 🤔

    I want to target wasm32 but the C ABI Convention get in the way.

    #rustlang #webassembly
    github.com/rust-lang/rust/issu

  11. I think I found a language that is suitable to write the wasm runtime for
    github.com/titzer/virgil

    I'm hoping that language support multi-values return and multi-memory since wizard-engine also written in Virgil 🤔
    github.com/titzer/wizard-engine

  12. I think I found a language that is suitable to write the wasm runtime for #nusa
    github.com/titzer/virgil

    I'm hoping that language support multi-values return and multi-memory since wizard-engine also written in Virgil 🤔
    github.com/titzer/wizard-engin

  13. devlog

    Let's dump my build graph in case I need to revisit again to build my own Rust crate for inline wasm text

    github.com/DrSensor/nusa/wiki/

  14. #nusa devlog

    Let's dump my build graph in case I need to revisit again to build my own Rust crate for inline wasm text

    github.com/DrSensor/nusa/wiki/

  15. devlog

    I think I can use this 😏
    doc.rust-lang.org/core/hint/fn

    (ugh, I think I should learn how to build my own allocator)

    Courtesy
    github.com/Speykious/cve-rs

  16. devlog

    dun..dun-dun
    Intent to Ship: WebAssembly Multi-Memory
    groups.google.com/a/chromium.o

    Oh yeah, multi-memory is coming!! It really scratch my head implementing Apache Arrow that share memory across instances and boundaries (i.e worker, shader, main-thread).

    But wait, I need a Rust crate/macro for wasm inlining (similar to `asm!()` or at least a `#[proc_macro] extern {}`). Using llvm-mc to write wasm specific stuff and link them with rust `output.{so,wasm}` really give me a headache.

  17. #nusa devlog

    dun..dun-dun
    Intent to Ship: WebAssembly Multi-Memory
    groups.google.com/a/chromium.o

    Oh yeah, #wasm multi-memory is coming!! It really scratch my head implementing Apache Arrow that share memory across instances and boundaries (i.e worker, shader, main-thread).

    But wait, I need a Rust crate/macro for wasm inlining (similar to `asm!()` or at least a `#[proc_macro] extern {}`). Using llvm-mc to write wasm specific stuff and link them with rust `output.{so,wasm}` really give me a headache.

  18. devlog

    Pomsky now have unit test (yay).
    pomsky-lang.org/blog/pomsky-0.
    I've been using code comments to annotate the success&fails criteria.
    Oh by the way "unit test" is a pun for "you need test"

  19. #nusa devlog

    Pomsky now have unit test (yay).
    pomsky-lang.org/blog/pomsky-0.
    I've been using code comments to annotate the success&fails criteria.
    Oh by the way "unit test" is a pun for "you need test"

  20. devlog
    github.com/llvm/llvm-project/i

    I guess I can finally enable debuginfo in wasm-opt after I compile my [component].wasm with rustc with debuginfo & DWARF enabled.

  21. #nusa devlog
    github.com/llvm/llvm-project/i

    I guess I can finally enable debuginfo in wasm-opt after I compile my [component].wasm with rustc with debuginfo & DWARF enabled.

  22. devlog

    Now I'm tempted to replace because separating type annotation into .d.ts didn't works on `class`and `abstract class`. Even `/ @abs​tract */ class`didn't exist.

    Thinking to replace it with because it has sane comment-based type annotation. But hold on, does it support Ruby style for type annotation?
    (I mean separation between file.js as implementation and file.d.ts as type annotation)

  23. #nusa devlog

    Now I'm tempted to replace #Typescript because separating type annotation into .d.ts didn't works on `class`and `abstract class`. Even #JSDoc `/ @abs​tract */ class`didn't exist.

    Thinking to replace it with #FlowType because it has sane comment-based type annotation. But hold on, does it support Ruby #RBS style for type annotation?
    (I mean separation between file.js as implementation and file.d.ts as type annotation)

  24. devlog

    Just finish replacing JSDoc with *.d.ts (inspired from Ruby ) so I can debug it easily without too much distraction.

    Somehow using either or make it difficult to reason when fixing many regression I encounter 😵‍💫

  25. #nusa devlog

    Just finish replacing JSDoc with *.d.ts (inspired from Ruby #RBS) so I can debug it easily without too much distraction.

    Somehow using either #JSDoc or #Typescript make it difficult to reason when fixing many regression I encounter 😵‍💫

  26. devlog

    LoL, I'm not sure why nix rebuild `pkgs.rome` instead of fetching it from cache.nixos.org 🤔

  27. #nusa devlog

    LoL, I'm not sure why nix rebuild `pkgs.rome` instead of fetching it from cache.nixos.org 🤔

  28. devlog
    Migrating my build system to xmake.lua

    I love this era of AI. Figuring out how to do this based on the official docs is impossible for me.

    phind.com/search?cache=ihtsqit

  29. #nusa devlog
    Migrating my build system to xmake.lua

    I love this era of AI. Figuring out how to do this based on the official docs is impossible for me.

    phind.com/search?cache=ihtsqit

  30. [side note for #nusa devlog]

    Now I just need something that can swap the runtime of #nixpkgs `nodePackages.rollup` (nodejs) with deno.

  31. #nusa devlog
    #rustwasm rant

    Somehow I decide to use #Rust rather than #Zig to implement Apache Arrow data type because Rust support multivalue return. However it can only return up to 2 value. And unlike Zig, Rust wasm can't emit nor import mutable-globals 😞. The `static mut` variables are stored in wasm linear memory instead of being `global mut`.

    If my guess is correct, `static mut` should be wrapped in `thread_local` macro to produce #wasm mutable-globals. But… it's not yet supported 🥲

  32. #webdev news

    auto accessor and stage-3 decorator landed to SWC 🥳
    github.com/swc-project/swc/iss

    I wish SWC have standalone CLI without requiring Node.js or whatever to make clean integration easier. I've been rewriting #nusa demo site in soupault + caddy + ncat + make and it's quite satisfying experience.

  33. #javascript news

    Resizable ArrayBuffer has been out for Nodejs and Deno. So the only one left behind now is Firefox :firefoxnew:

    Maybe I should consider to use Chrome as my default browser for developing #nusa 😂

    developer.mozilla.org/en-US/do

  34. #nusa #webassembly devlog

    Didn't expected that #wasm GC already in origin trial. Seems I need try if Host Types in wasm GC for JS API work in my case before moving to #ApacheArrow.
    github.com/DrSensor/nusa/issue

    I wonder if there's a language or compiler that already support wasm GC 🤔 (I doubt rust js_sys and wasm_bindgen already support it)

  35. #nusa devlog

    Yes, `Symbol.dispose` is what I need to specify the destructor for handling page transition in my framework.
    docs.google.com/presentation/d

  36. #nusa devlog

    thinking about mode of reactivity inside effect/compute 🤔

  37. Tips for simple ultra fast testing:

    - use test library (not runner) like zora, tape, uvu, or jasmine-core for testing dynamic page/content (i.e SPA, Island)

    - use DOM emulation like linkedom or happy-dom for testing static page/content (i.e SSG, SSR)

    #nusa #webdev #TDD

  38. #nusa devlog

    I wish #typescript have ability to define the type of imported variable on specific scope 🤔