Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
vshymanskyy
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
31.
▲
by
vshymanskyy
3y ago
It wasnt
32.
▲
by
vshymanskyy
3y ago
OP here, ready to answer questions
33.
▲
JIDL – A simple, flexible, JSON-based Interface Definition Language
(github.com)
5 points
by
vshymanskyy
3y ago
|
0 comments
34.
▲
by
vshymanskyy
4y ago
UTF8 encoding is stable in that aspect for decades, so don't expect any changes soon (or ever). Actually, gaps even increased (5-byte and 6-byte code points were removed) when they realized that UTF-8 doesn't need that much of enc
35.
▲
MUON: Compact and simple binary format, uses gaps in Unicode encoding for markup
(docs.google.com)
8 points
by
vshymanskyy
4y ago
|
3 comments
36.
▲
by
vshymanskyy
5y ago
Just compiled a list: https://github.com/vshymanskyy/StandWithUkraine/blob/main/do...
37.
▲
Run native, statically-compiled apps on any embedded platform using WebAssembly
(twitter.com)
2 points
by
vshymanskyy
5y ago
|
0 comments
38.
▲
Each pixel on this board is controlled by a different WebAssembly app [video]
(twitter.com)
1 points
by
vshymanskyy
5y ago
|
0 comments
39.
▲
Wasm3 release v0.5.0: Multi-Values, Bulk Memory ops, Apple M1 support
(github.com)
6 points
by
vshymanskyy
5y ago
|
0 comments
40.
▲
Node-WASM-run allows running arbitrary WASM/WASI files in V8 engine
(twitter.com)
1 points
by
vshymanskyy
5y ago
|
0 comments
41.
▲
A curated list of language-agnostic WebAssembly tools
(github.com)
1 points
by
vshymanskyy
6y ago
|
0 comments
42.
▲
by
vshymanskyy
6y ago
GitHub release: https://github.com/wasm3/wasm3/releases/tag/v0.4.9
43.
▲
Wasm3 v0.4.9: seamless WebAssembly/WASI tracing, Python integration, etc.
(wapm.io)
3 points
by
vshymanskyy
6y ago
|
1 comments
44.
▲
by
vshymanskyy
6y ago
Fixed the description, thanks!
45.
▲
by
vshymanskyy
6y ago
The question is too generic. Wasm3 has some quirks when used directly on micro-controllers, and this is exactly why we're doing Wasm3 - to find if/how we can use it. For example, Linux uses eBPF for similar purposes. PAF could be
46.
▲
by
vshymanskyy
6y ago
We have some examples using pygame for image output. But it's not cosmopolitan-related. Check for pywasm3 python module.
47.
▲
by
vshymanskyy
6y ago
But it requires Ahead-of-Time compilation, as well as full-fledge optimizing compiler. But for sure, this approach has it's own use cases.
48.
▲
by
vshymanskyy
6y ago
Author here. For Apple M1, I'd suggest building a native build of wasm3, to get full speed of "the fastest interpreter".
49.
▲
by
vshymanskyy
6y ago
Author here. In Wasm3, we have implemented some workarounds for that. We can easily run on Nordic nRF51 with just 16 KB of RAM!
50.
▲
Full WebAssembly and WASI self-hosting finally achieved
(twitter.com)
3 points
by
vshymanskyy
6y ago
|
0 comments
51.
▲
by
vshymanskyy
6y ago
It works on FreeBSD, at the end of the day: https://twitter.com/wasm3_engine/status/1253707153619247106
52.
▲
Wasm3 runs WebAssembly on Big-Endian systems
(github.com)
3 points
by
vshymanskyy
6y ago
|
3 comments
53.
▲
by
vshymanskyy
7y ago
It might be. RISC-V is looking into this as well. In meantime, you can try Wasm3 interpreter, which can run wasm bytecode on all sorts of things. Check out this demo: https://github.com/vshymanskyy/Wasm3_RGB_Lamp
54.
▲
Show HN: DIY Animated RGB Lamp, Runs WebAssembly
(github.com)
4 points
by
vshymanskyy
7y ago
|
0 comments
55.
▲
Show HN: TinyGo interoperability with Arduino-based WebAssembly runtime
(twitter.com)
3 points
by
vshymanskyy
7y ago
|
0 comments
56.
▲
by
vshymanskyy
7y ago
V8 has a built-in (pure, no JIT etc.) interpreter of WASM. Which is quite slow according to this test.
57.
▲
by
vshymanskyy
7y ago
Yup. TinyBLE is nRF51 SoC with 16Kb SRAM as well.
58.
▲
by
vshymanskyy
7y ago
Yeah, but... It's not only the "threaded code" approach, that makes Wasm3 so fast. In fact, Intel's WAMR also utilizes this method, yet is 30 times slower..
59.
▲
by
vshymanskyy
7y ago
It's not only the "threaded code" approach, that makes Wasm3 so fast. In fact, Intel's WAMR also utilizes this method, yet is 30 times slower..
60.
▲
by
vshymanskyy
7y ago
It's not that bad even in debug mode (or without TCO). Just not optimal. Also, there is a way to rework this part, so it does not rely on compiler TCO.
More ›