exception-handling VS standards-positions

Interesting article, thanks!

Notes on the issues mentioned there:

* The need for a manual shadow stack: This is fixed in WasmGC (in the same way it works in JS, as the link mentions).

* Lack of try-catch: This is fixed by the Wasm exception handling proposal, which has already shipped in browsers, https://github.com/WebAssembly/exception-handling/blob/main/...

* Null checks: Mostly fixed by WasmGC. The spec defines non-nullable local types, and VMs can use the techniques the article mentions to optimize them using signals (Wizard does, for example).

* Class initialization: This is a difficult problem, as the article says. J2Wasm and Binaryen are working to optimize it through static analysis at the toolchain level. Here is a recent PR I wrote that makes progress there: https://github.com/WebAssembly/binaryen/pull/6061

* The vtable overhead issue the article mentions may be a problem. I'm not aware of good measurements on it, through. There are some ideas on post-MVP solutions for method dispatch that might help, but nothing concrete yet.

* Checks for null and trapping: There has been discussion of variants on the GC instructions that throw instead of trap. Measurements, however, have not shown it to be a big problem atm, so it is low priority.

The author is right that stack walking, signals, and memory control are important areas that could help here.

Overall with WasmGC and exceptions we are in a pretty good place for Java as emitted by J2Wasm today: it is usually faster than J2CL which compiles Java to JavaScript. But there is definitely room for improvement.

It uses https://github.com/WebAssembly/exception-handling/blob/master/proposals/exception-handling/Exceptions.md which is actually supported by all major browsers already.

Yep, you're right. It's also more than just the DOM, it's web APIs in general, such as fetch, audio, webgl/webgpu, etc. WASM still needs GC, exceptions, and WASI to be able to fully interop with any host without any of the current limitations. This'll take a few years. I'm looking forward to the future in which I will be shipping WASM-only web apps to my users.

Its a part of the wasm plan to support gc https://github.com/WebAssembly/gc exceptions https://github.com/WebAssembly/exception-handling

As of Node.js 17, the v8 JavaScript engine has been updated to v9.5. The changes in this release are primarily aimed at expanding internationalization for dates and calendars as well as for the output of time zones. It also implements the WebAssembly Exception Handling proposal, designed to reduce overhead compared to current JavaScript-based workarounds.

Agreed, although at some point in a not very far feature most of those missing features will resolved. So in my mind is just a matter of time. The Wasm Community group is doing an awesome work on that :)

Here are a few examples of what needs move forward in Wasm:

* [1] Wasm Exceptions Handling: Right now Wasm is missing a way to handle exceptions natively (C++ programs can only compile to Wasm using the asyncify or longjmp/setjmp tricks via Js try/catch)

* [2] Wasm GC: Wasm Binary files are quite big (specially in interpreted languages). This is partially caused by the GC being included in the Binary itself. The GC proposal will solve this while also providing faster execution.

* [3] Wasm 64-bit Memory: currently Wasm can only operate with 32-bit data. In some contexts you may want you operate with more than 4GB of memory (for example, when operating over terabytes of data). The 64-bit memory proposal will solve that.

[1]: https://github.com/WebAssembly/exception-handling

You can read through the conversations to understand more of the context

https://github.com/mozilla/standards-positions/issues/100#is...

https://github.com/mozilla/standards-positions/issues/95#iss...

https://github.com/mozilla/standards-positions/issues/336

The main struggle is around giving informed consent that explains the risks. Understandably, browsers don't want to ship a "Set my printer on fire" button.

You can check why Mozilla and Apple have opted to not support this.

https://github.com/mozilla/standards-positions/issues/154

https://github.com/WebKit/standards-positions/issues/28

Neither Mozilla or Webkit are satisfied that the proposal is safe by default, and contains footguns for the user that can be pretty destructive.

FWIW Mozilla updated their position on Web Serial API to "neutral" and clarified that they might be okay with enabling the API with an add-on.

https://mozilla.github.io/standards-positions/#webserial

Allowing serial but not HID would be really strange. With HID you get standard identifiers that let you filter out devices that are too dangerous for the web. With serial you get nothing. Even if you know a device is dangerous, there's no way to protect users from it.

Hasn't FireFox been dragging their asses on @scope? https://github.com/mozilla/standards-positions/issues/472

It took years to just convince them of the need for it. And I'm not sure anyone got convinced vs Chrome had already shipped it and Safari has it planned so they caved in.

Hard to believe FireFox used to be a leader of the modern web.

As mentioned by others, OK idea, but not a fan that this isn't standardized. After a quick search+peruse, these seem to indicate that it's not around the corner either. Happy (/hope) to be corrected.

https://github.com/whatwg/html/issues/4180

https://github.com/mozilla/standards-positions/issues/990

Mozilla's position on these specs is nicely outlined publicly and transparently as part of their standards-positions project: https://github.com/mozilla/standards-positions/issues/100

I'm kinda glad it's not implemented in my browser, to be honest, because the whole thing seems like a security nightmare.

It's a shame it impacts some hobby usecases, but I don't think this outweighs the reasoning set out on the GitHub issue.

This should have big warnings on it. Some of these are not web standards; they are features implemented unilaterally by Google in Blink that have been explicitly rejected by both Mozilla and Apple on privacy and security grounds.

Take Web Bluetooth, for example:

Mozilla:

> This model is unsustainable and presents a significant risk to users and their devices.

— https://mozilla.github.io/standards-positions/#web-bluetooth

Apple:

> Here are some examples of features we have decided to not yet implement due to fingerprinting, security, and other concerns, and where we do not yet see a path to resolving those concerns

— https://webkit.org/tracking-prevention/

This is Microsoft’s Embrace, Extend, and Extinguish bullshit applied to the web platform by Google. Google keeps implementing these things despite all other major rendering engines rejecting them, convinces people that they are part of the web, resulting in sites like this, then people start asking why Firefox and Safari are “missing functionality”. These are not part of the web platform, they are Google APIs that have been explicitly rejected.

Is BLE a PWA requirement? I think they explained their position pretty well here, regardless of whether I agree:

https://github.com/mozilla/standards-positions/issues/95#iss...

I took a glance at Can I Use what the difference between the last public release of Firefox and Chrome is [1] and they don't really have that big of a difference in the eyes of normal use-cases? Some of these aren't implemented purely because of privacy reasons, the proposals aren't finished yet or complexity [2].

Why would Firefox need to change to Chromium engine? The only websites I notice that don't work with Firefox is because of user-agent targetting or just putting 5-second time-outs in Youtube code on non-chrome webbrowsers [3].

Can you give some examples of websites not working on Firefox?

[1] https://caniuse.com/?compare=chrome+120%2Cfirefox+121&compar...

[2] https://mozilla.github.io/standards-positions/

[3] https://www.neowin.net/news/youtube-seemingly-intentionally-...