Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
azakai
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
181.
▲
by
azakai
3y ago
I like the term "VMless" (or "VM-less") for this. (Inspired by "serverless" I guess)
182.
▲
by
azakai
3y ago
That's true but it's a separate matter. Yes, vegetable and seed oils can be unhealthy if they are processed into trans fats, but otherwise they can be healthy. Whole Foods, for example, banned trans fats 20 years ago, but they s
183.
▲
by
azakai
3y ago
To add to that, here is the relevant part of the article: > We hired the law firm Buzko Krasnov to file a pre-trial claim against Apple.
184.
▲
by
azakai
3y ago
At the very end of the blogpost the author asks why not compile C# to JavaScript, like F# (Fable) does? The author thinks that would be the best solution overall, and is surprised it has not happened yet. In fact that has happened, see JSIL
185.
▲
by
azakai
3y ago
Yes, there is a big difference between having access to the weights and code and having a license to use them in different ways. It seems obvious Microsoft has a license to use them in Microsoft's own products. Microsoft said so direct
186.
▲
by
azakai
3y ago
This sounds like very useful protection to me: yes, you can have crashes in prod, but those are unexploitable at least. That's a very big deal in some systems.
187.
▲
by
azakai
3y ago
You can compile the game to wasm and run it in a web environment. That can be a wrapper around a system webview, so you don't need to actually bundle a web engine in your app. The web environment would provide all the API support for y
188.
▲
by
azakai
3y ago
The Binaryen wasm optimizer (mentioned in the article) is always open for contributions! https://github.com/WebAssembly/binaryen
189.
▲
by
azakai
3y ago
.NET's runtime gives it capabilities like finalizers, JITing, debugging, etc. - it's very comprehensive. J2Wasm (Java on WasmGC) is making do without those for now, and just shipping minimal support for a large/useful subset
190.
▲
by
azakai
3y ago
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
191.
▲
by
azakai
3y ago
For now I think that's true. But there are plans to add features like interior pointers to WasmGC, that could allow a lot more eventually: https://github.com/WebAssembly/gc/blob/main/proposals/g
192.
▲
by
azakai
3y ago
> Would it make sense for the runtime to expose an inbuilt alloc library that C/C++/Rust can use? A difficulty with that approach is that the alloc library's behavior would need to be standardized and deterministic. If it
193.
▲
by
azakai
3y ago
Yes, that's definitely true: a single GC will not be optimal for everything, or even possible. Atm interior pointers are not supported at all, for example, but they are on the roadmap for later: https://github.com/WebAs
194.
▲
by
azakai
3y ago
> Any non-toy uses of WebAssembly are gonna be new developments. Major uses of WebAssembly include things like Photoshop, Unity, and many other large existing codebases.
195.
▲
by
azakai
3y ago
Perhaps another factor the article was referring to is that Go compiles ahead of time to static binaries. Java is typically JIT compiled which means extra overhead and unpredictability at runtime as code warms up.
196.
▲
by
azakai
3y ago
It should certainly lower the bar of adopting a new codec if the implementation is in a memory-safe language. Even so, it is more code, and somewhat more risk. Lack of safety elsewhere might end up using code that is otherwise safe in order
197.
▲
by
azakai
3y ago
An example of a recent compile target that breaks on unaligned pointer accesses was asm.js. There, a 32-bit read turns into a read from a JavaScript Int32Array like this: HEAP32[ptr >> 2] The k-th index in the array contains 4 bytes o
198.
▲
by
azakai
3y ago
This makes me suspect there is even more to this (amazing!) story: > Their saga is somewhat incomprehensible, not least because it’s eventually conveyed in two separate movies, intended to be screened either by alternating scenes from on
199.
▲
by
azakai
3y ago
My experience is exactly the opposite. Most of my social sphere has already moved to Bluesky. Also most of the top posters I used to follow on Twitter have either migrated or post on both. As a result I've been spending more time on Bl
200.
▲
by
azakai
3y ago
There is also Type-After-Type: https://dl.acm.org/doi/10.1145/3274694.3274705 (though maybe that's covered by what the author meant by "arenas").
201.
▲
by
azakai
3y ago
The Chinese Room Argument is considered by many to be incorrect on those matters. But it has led to decades of interesting philosophical debate, which is worth reading to see the various perspectives. This is an excellent summary: https:&#
202.
▲
by
azakai
3y ago
Yes, and that's one of my favorite Chomsky points. I don't remember the exact phrasing, but something like: "Language is innate in humans because in every household practically all children learn it while none of the pets do.
203.
▲
by
azakai
3y ago
75,000 words is a drop in the bucket for A Song of Ice and Fire: https://blog.fostergrant.co.uk/2017/08/03/word-counts-popula...
204.
▲
by
azakai
3y ago
But see the section "Intentional differences from Python", which is after your quote, https://docs.modular.com/mojo/why-mojo.html#intentional-diff... > our approach to compatibility is two fold: > 1. We
205.
▲
by
azakai
4y ago
That is not true for compiling to wasm. You can use upstream LLVM (and Emscripten, which uses it) with -flto normally on all platforms. It is basically cross-compiling, so it does not matter at all on what platform you are building, and Win
206.
▲
by
azakai
4y ago
Yes, exactly. LLVM docs: https://llvm.org/docs/LinkTimeOptimization.html
207.
▲
by
azakai
4y ago
> Cheerp directly uses the LLVM bytecode format as the intermediate representation, for both object files and libraries. The paragraph with this might be a little confusing. It suggests that this is a benefit of Cheerp, but upstream LLVM
208.
▲
by
azakai
4y ago
There are many other factors that could be relevant here. As the article says, ultra-processed food tends to be low in nutrients. Given the little we know, that seems as likely a cause as anything else here. Also, these are not randomized s
209.
▲
by
azakai
4y ago
> What does the edge mean? I was confused by that too. I've seen "edge" mean certain things in the web space (Cloudflare workers, Fastly compute, etc.), but in this article it apparently means "IoT" (Internet of
210.
▲
by
azakai
4y ago
I read this book two years ago. It's incredibly interesting, and made me look at the world differently. Highly recommended.
More ›