Khaidi Chu

libuv 本来就有多线程的 API 呀。

你指的是 multi-reactor 模式，这个可以把 Issue 提到 libuv 的仓库下。

I think we need to inherit a class from `sf::Shape` a high degree of freedom and expose it to JavaScript side. And make sure that JavaScript can "inherit" that exposed...

Due to the V8 bug is not public (https://bugs.chromium.org/p/chromium/issues/detail?id=1394403), I think we should cherry-pick this PR to any influenced versions of Node.js, not only for v18.x.

> Despite the performance boost, I'm generally -1 on introducing a new dependency and a different codepath for base64 encoding/decoding only for atob/btoa. If anything, we should be looking at...

@addaleax @jasnell I've updated, removing modp_b64 and using `base64_encode`. And got the new benchmark: ```bash $ ./node benchmark/run.js --filter buffer-btoa.js buffers buffers/buffer-btoa.js buffers/buffer-btoa.js n=32 len=67108864: 3.3501380480392817 $ ./node benchmark/run.js --filter...

> @XadillaX In order to merge this, I think we should: > > * Remove the C++-backed atob and btoa implementations > * Remove their benchmarks > * Undo the...

@addaleax I've tried to use modp_b64 to instead of current and found the performance is almost the same (https://github.com/nodejs/node/compare/c975dff3c0f0f1ecb1574f3b10dd1d135a7704db...XadillaX:base64-perf). So the way to improve this PR is just to make...

``` javascript function pascal(depth) { return (new Array(depth - 1)).join("€").split("€").reduce(function(arr) { return arr.push((new Array(arr.length + 1)).join("囍").split("囍").map(function(_, idx) { return idx || idx !== arr[arr.length - 1].length ? arr[arr.length - 1][idx...

> I don't think this should exist, it's basically "run at the end of the microtask queue after the next tick", which is not an idiom anyone should be reaching...