News for hackers. Stuff that matters.
Re:What asm.js is and isn't by lmm on Friday June 21, 2013 @09:48AM (#5917987)
So Mozilla now believes other browser makers can and will optimize "generic" javascript well enough to match the performance of their own specialized engine, without implementing any specializations: "To say that asm.js code will be "unusably slow" on browsers other than Firefox implies that those other VM devs can't match a level of performance that was shown to be possible by their peers. That is a ridiculous thing to say given how talented those devs are, which has been proven countless times."

But two years ago Mozilla's position was "A Dart to JS compiler will never be "decent" compared to having the Dart VM in the browser."

So what's changed?

Re:What asm.js is and isn't by azakai on Friday June 21, 2013 @01:25PM (#5919515)
Got a link to that quote? I don't think I've seen it, interested to see the context.

Anyhow, a lot has changed in two years. dart2js now produces JS that, at least on the benchmarks on the Dart perf site, compare well with handwritten JS. This was not obviously possible two years ago! In fact back then I would also have been extremely skeptical that C++ compiled to JS would get close to native speed as well, but it has.

Re:What asm.js is and isn't by lmm on Friday June 21, 2013 @01:40PM (#5919649)
It was from here; https://news.ycombinator.com/item?id=2982256 . Possibly not the best example but I was looking for something I could find quickly that was representative of the mozilla position as I remembered it.
Re:What asm.js is and isn't by azakai on Friday June 21, 2013 @02:09PM (#5919875)
Not sure where to look, there is a lot of stuff there: In the HN comments? The linked-to blogpost? Or comments in the linked-to blogpost? I skimmed them quickly, but didn't see what you are quoting yet.