cglue VS interface-types

Shameless plug - you may want to check out cglue, which lets you mix and match rust versions, and even implement plugins in other languages like C++: https://github.com/h33p/cglue

You can grab CGlue from my GitHub.

I honestly wouldn't say that much has changed during this year. I do see some progress on new libraries like xlang_abi, safer_ffi or cglue, but it's still too early to know, and existing ones like abi_stable move quite slowly.

You might wanna take a look at cglue: https://github.com/h33p/cglue

Importing Rust is the same process as you'd use to import a C library, but there's a lot of people actively invested in this area so there's likely already automated tooling of some sort. Google uses https://cxx.rs for automating C++ bindings to Rust libraries, and lately there's been experiments like cglue to automate C bindings.

I'm proud to announce one final CGlue release for this year, the highlight being ability to use layout_checks feature on non-git builds!

I have just added pre-generated headers, they are now on the repo. Yup, certainly not the cleanest, some whitespaces need fixing, but it wasn't the highest priority as it didn't affect functionality.

Depending on how big your API surface is (and how trait heavy it is), you may want to look into [shameless_plug] automatically generating it with cglue. That's what we did for memflow. [/shameless_plug] It's goals are not much different from abi_stable (in fact, rather a small subset of it focused on traits), but does create C-compatible code.

It seems like we are trying to solve roughly the same problems when it comes to loading plugins and keeping it all ABI safe. Might be worth checking out CGlue (full disclosure: it's my project). While I do not have async trait support, it would be a really interesting challenge to tackle. The ultimate goal for the project is to be a dead simple code generator that would even allow to write plugins in C/C++, and I currently have sweeping changes underway to get it there.

Some things that might greatly increase wasm usage and overall tooling:

1) Tools that run docker containers and serverless function services (like AWS lambda) to support providing a .wasm files instead

2) Garbage collection in the runtime to make GC languages easier to port to wasm

3) Dynamically typed languages (NodeJS, Python, Ruby) being able to compile to webassembly directly instead of porting the runtime to webassembly and then running the code through the runtime. This is a big ask though, basically needs to redesign the runtime completely

4) wasm-DOM bindings will enable other languages to do HTML rendering which will require new web frameworks for every language that wants to take over the space from JS. This will lead to (even more) fragmentation of the web ecosystem

5) A new wasm-first SDK (unrelated to the DOM) for building cross platform applications. I can see this taking off only if it is built-into the browsers and backed by some standards committee, so not very likely I think

6) Something like the Interface Types proposal ( https://github.com/WebAssembly/interface-types/blob/main/pro... ) becomes a thing allowing wasm programs to be consisted of modules written in several different languages and being able to call said modules with low or 0 runtime performance hit (and of course, no compilation to multiple CPU archs). So much of programming ecosystems are locked to specific languages (like data science with python) when there is little technical reason for it be like that.

AFAIK GC is irrelevant for "direct DOM access", you would rather want to hop into the following rabbit hole:

- reference types: https://github.com/WebAssembly/reference-types/blob/master/p...

- interface types (inactive): https://github.com/WebAssembly/interface-types/blob/main/pro...

- component model: https://github.com/WebAssembly/component-model

If this looks like a mess, that's because it is. Compared to that, the current solution to go through a Javascript shim doesn't look too bad IMHO.

I'm not really optimistic that particular aspect will get much improvement. Many people expected interface types to come save the day, but after a looong stagnation that proposal has been archived (for now) in favour of component types, which has much less potential for performance gains.

Really good questions. Unfortunately, most of the issues I found back then were fundamental ones. I've seen that Wasm has deprecated "Interface Types" and is now working on the "Component Model". But even then, as far as I understand that would only avoid the serialization and deserialization steps, and you would still need to copy complex types. It will be more performant, but I don't think it would be enough for Tremor either.

wasm doesn't know anything about the outside world on purpose. This allows it to be used in other domains. For direct access to the DOM et al, interface types are being developed. It's a non-trivial problem to interoperate with a dynamically typed GC'd language from any statically typed no-GC language that can compile to wasm.

You may want to look into WASM interface types, which is defining what amounts to am IDL for WASM and different languages have common calling conventions: https://hacks.mozilla.org/2019/08/webassembly-interface-type...

I don’t know if there’s a better intro article. I believe this is the current iteration of the proposal: https://github.com/WebAssembly/interface-types/blob/main/pro...

Yes, and if I'm not mistaken, this is the RFC

I see no reason why not. See the interface types proposal for a proposed solution.

Wanted to get thoughts from the Rust experts on this - the author of the Yew framework seems to think that Web Assembly Interface Types (https://github.com/WebAssembly/interface-types/blob/master/proposals/interface-types/Explainer.md) will allow Yew to eventually become faster than Vue, React, Angular, etc. Is there general consensus on this in the Rust community? The prospect of mixing Rust (for the performance critical pieces) with TS on the front end doesn't seem super appealing to me.