Announcement

Collapse
No announcement yet.

WASMtime 0.12 Released For The JIT-Style WebAssembly Runtime

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • qarium
    replied
    Originally posted by Michael View Post

    I tried out v0.12 but didn't get too far... Was compiling some basic C benchmarks to WASM and also trying some other .wasm files to try with wasmtime. Unfortunately none of what I tried worked with wasmtime, as I was hoping to add wasmtime as another CPU benchmark but looks to be too early.
    can we have PTS in a browser WebAssembly Edition?
    can you port PTS to WASM/WASI ?

    I think web browser based PTS would be great to a big market and the usability would be great for non-technical people.

    I really will send you paypal money if you start to go in this direction. I think other people should sponsor you to if they want Web Browser editiion of PTS

    Leave a comment:


  • starshipeleven
    replied
    Originally posted by kpedersen View Post
    I like this idea but unfortunately I am pretty sure that the wasi standard and implementation will have changed by then rendering our isolated "blobs" un-runnable.

    To see what I mean, try executing some Java bytecode compiled with an early (pre 1.2) javac. It will not run within any modern VM.
    Java is a fucking trainwreck that failed to reach more or less all it promised.

    You can run your application with any Java runtime! Nope, sometimes it will break horribly if you update the runtime beyond version X without fixing the source code.

    You can run it on any arch! Nope, sometimes it will break horribly outside of x86 for some unknown reason you can't figure out without wasting weeks on it, and in most cases your developers aren't low-level enough to actually be able to debug and understand the issue (not have anywhere near the time to do so)

    I don't see why WASM should automatically have the same issues, retrocompatibility is a choice, and Java development never provisioned for that at all (just assumed that old stuff would work later).

    Leave a comment:


  • kpedersen
    replied
    Originally posted by oleid View Post
    So if you add wasi support for you project today, you might be able to easily update the sandbox's content in ten years using c++30.
    I like this idea but unfortunately I am pretty sure that the wasi standard and implementation will have changed by then rendering our isolated "blobs" un-runnable.

    To see what I mean, try executing some Java bytecode compiled with an early (pre 1.2) javac. It will not run within any modern VM.

    Perhaps isolate your components in really *really* dumb C89 and you probably have a better chance of it still compiling and working in 10 years :/

    Leave a comment:


  • Michael
    replied
    Originally posted by andrei_me View Post

    Just to be clear, those wasm files that didn't work with WASMtime but we're working in the browsers, right?
    Yes at least some of them (some of the .wasm tests used in a few of my browser tests) while others were fresh built from wasmer and other Rust scripts.

    Leave a comment:


  • oleid
    replied
    I know exactly what I would use that for : same as Mozilla : isolate independent parts of an application, being forced to define a sane IO layer and being able to easily test that stuff with a scripting language like python. Also, you'll get a stable compile target. So if you add wasi support for you project today, you might be able to easily update the sandbox's content in ten years using c++30.

    Of course there is no need for wasi on most of the cases, but for sandboxing and stable target.

    Leave a comment:


  • Ironmask
    replied
    Wow, I didn't know this project had bindings for other languages like C#, this is really great, they're really aiming for a practical universal WASM runtime. I don't know what I'll use this for, but I'll definitely be looking out for an excuse and weigh my options.

    Leave a comment:


  • andrei_me
    replied
    Originally posted by Michael View Post

    I tried out v0.12 but didn't get too far... Was compiling some basic C benchmarks to WASM and also trying some other .wasm files to try with wasmtime. Unfortunately none of what I tried worked with wasmtime, as I was hoping to add wasmtime as another CPU benchmark but looks to be too early.
    Just to be clear, those wasm files that didn't work with WASMtime but we're working in the browsers, right?

    Leave a comment:


  • Michael
    replied
    Originally posted by CuriousTommy View Post
    Seem pretty interesting, I am going to give it a try later when I get some free time.
    I tried out v0.12 but didn't get too far... Was compiling some basic C benchmarks to WASM and also trying some other .wasm files to try with wasmtime. Unfortunately none of what I tried worked with wasmtime, as I was hoping to add wasmtime as another CPU benchmark but looks to be too early.

    Leave a comment:


  • CuriousTommy
    replied
    Seem pretty interesting, I am going to give it a try later when I get some free time.

    Leave a comment:


  • WASMtime 0.12 Released For The JIT-Style WebAssembly Runtime

    Phoronix: WASMtime 0.12 Released For The JIT-Style WebAssembly Runtime

    Announced last November was the Bytecode Alliance with a goal of running WebAssembly everywhere. This effort by Intel, Red Hat, Mozilla, and others has resulted in a new release today of wasmtime, their JIT-style runtime for WebAssembly on the desktop...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite
Working...
X