Nice, that might make it almost as fast as Chrome!
Jokes aside, this is garbage clickbait - the reality as stated in the linked post is much more mundane: "Some processing tasks are now more than 75 times faster in Firefox"
Optimizing the Ion compiler back end
41–45 of 45 posts
Re: Optimizing the Ion compiler back end
#42From a glance, they replaced linked lists with vectors, changed dominator tree building to semi-NCA, and used sparse bitsets to reduce memory usage.
Very niche but effective. When people ask why do FAANGs focus on data structure algos during interviews, it's for cases like this. With that said, I understand candidates frustration since most end up working on higher level network services.
Re: Optimizing the Ion compiler back end
#43Ok, but have they fixed their awful javascript performance?
Re: Optimizing the Ion compiler back end
#44I welcome this and look forward to seeing the improvements. At the moment I am running some TSS code compiled down to WASM. The last time I executed DKG and signing Webkit took ~25s, Chrome took about ~39s and Firefox took nearly 4 minutes so there is clearly lots of room for WASM optimization in Firefox!
Re: Optimizing the Ion compiler back end
#45Is there any reason to think this is anything but great?
> Is there any reason to think this is anything but great?
like in 75X great, or what ? /s