JavaScripts in the JavaScripts [pdf]

It's quite hard to follow without the associated talk.

I do recall past threads where JavaScript natives were rewritten in JavaScript and shown to be faster, then they were shown to be non-standard because of a heap of weird edge cases that have to be supported correctly. It hadn't yet been shown that a fully standards compliant implementation would be faster.

I think a huge amount rests on slide 8 "No JS/C++ transition cost". It should probably be the subject of significant analysis on it's own.

The calulation is min(Ctime + TransitionCost, JStime). Knowing all of these values individually would give a better picture. Is it better to Minimize the transition cost or eliminate it? It depends on how much less Ctime is compared to JStime.

Here's a link to the author's blog post: http://wingolog.org/archives/2014/11/09/ffconf-2014

I really got into this stuff a while ago, even fixing a few bugs in Continuum, an awesome (but fairly buggy) JS-in-JS implementation by Brandon Benvie: http://benvie.github.io/continuum/.

For those who like this kind of thing, I compiled a little list of JavaScript libraries for working with JavaScript: http://notes.richdougherty.com/2014/03/javascript-language-l.... Let me know if you have any more suggestions!

I stopped reading when I got to the part about js being faster than C++.

Is there a video of the presentation?