Hacker News new | ask | show | jobs
by lurenjia 3318 days ago
Major browser engines all have about 20 years of development history. Web specs are ever growing and piled higher and higher. Major web browsers have large engineering team and lot resources. It's unlikely Servo can catch up.
1 comments
lelandbatey 3318 days ago
Servo is already a runnable browser, and I'm amazed how fast it is. It still has rendering issues with plenty (any reasonably complicated media-heavy site, i.e.: cnn.com) of sites, but even then, it's already an amazing piece of technology. On top of that, the point of Servo isn't to be a brand new full on browser, it's to be a proving ground for a next-gen engine.

And it's proving to be fast, safe, and the future of browsers.

link
markdog12 3317 days ago
> Servo is already a runnable browser, and I'm amazed how fast it is

First of all, I think the idea behind Servo is awesome, and I follow it. But I've been testing it on Mac OS and Windows, and it is not a runnable browser, nor fast (as expected!). CPU is often fully pegged and it's very iffy if any UI elements or page loads work. Not to say they won't get there, but it's still very, very early and buggy.

link
Manishearth 3317 days ago
Yeah, we don't track perf regressions like released browsers so often we land something that negates the performance benefit. The current servo releases are just alpha, so it's not too important to stay on top of, but we should probably start caring about this more.

We had a similar issue with Stylo (Servo style system in gecko) recently where there were bugs in the parallelism code making us slower than gecko. Fixed, now we're faster again. We only recently started tracking performance properly, and it was caught and fixed in a few weeks.

link