The current WHATWG URL parser implementation is written in C/C++ and incurs a fairly significant cost crossing the JS/C++ boundary. It should be possible to realize a significant performance improvement by porting the implementation to WASM (similar to how the https://github.com/nodejs/undici project has seen a massive performance boost out of moving llhttp parser to wasm).

If someone wanted to

I've ran into this issue for a couple hours and I ended up editing the dist library adding two new functions called fetchVideo and bufferToVideo that works pretty much like the fetchImage and bufferToImage functions.

I'll leave it here to help somebody else with the same issue and in case someone wants to include it on future releases.

face-api.js

...
exports.fetchVideo = fetc

This would be a nice little feature to have, Excel users are using this a lot.

Currently, when entering epic mode the README is frozen in the last level of the tower. When you're trying to fine-tune the score for a level other than the last one, it would be helpful if we had the README for that level available. The proposal is that when entering epic mode, the README is updated with all levels, one following the other.

Example:

# Starbolt - beginner

• Source

JsPointerToString(NULL, 0, &str);

• Expected
  str = empty string

• Actual
  failed and returns JsErrorNullArgument

I think this should be possible

Got has a very good coverage test-wise. Would be nice to port relevant tests that we're missing over to Ky.

Note: Not for https://github.com/sindresorhus/ky/blob/master/test/browser.js, but for everything else.

System information

• node version: v10.13.0
• npm or yarn version: 6.4.1
• OS/version/architecture: gLinux 64-bit (Google-internal, like Debian Testing)
• Applicable nodegit version: v0.23.0 built from source

https://www.nodegit.org/api/tree/ has an example of how to use tree.walk(), but it doesn't mention a critical detail, which is mentioned in an example:
https://github.com

I'm not sure if the way this behavior is handled is intentional or not, but in MRI Ruby a TypeError is immediately thrown when trying to call send with a nil method name.

class Foo
  class << self
    def method_missing(name, *args, &block)
      puts "calling method missing"
      
      super
    end
  end
end

Foo.send(nil)

In MRI Ruby that puts statement will neve