4 ms·

>> "This is not the problem. You can easily use i.e. React and some related libs on mobile without slowing down the experience at all." Hasn't Facebook proven

>> "This is not the problem. You can easily use i.e. React and some related libs on mobile without slowing down the experience at all." Hasn't Facebook proven this to be incorrect. Facebook mobile app was originally just HTML5 & JavaScript. The experience was so slow, they ditched it and instead built a native iOS & Android app. [1] [1] https://techcrunch.com/2012/12/13/facebook-android-faster/ https://techcrunch.com/2012/12/13/facebook-android-faster/

manmal 10y ago
Two more reasons why this comparison is not on point: - On iOS, the FB App had to make do with a really slow JS engine at the time. IIRC Safari's JS performance was x5 compared to JS performance in apps' web views? This has more to do with how iOS treats apps, rather than how fast web pages in Safari are or were. - The FB apps were okayish at the time, given that they were web based. If you manage to get your web app to that level, your site is ok. Now add to that the performance gains and browser optimizations of the last 4 years and you get a system that flies. Native apps are still faster and respond quicker to touch, but you cannot hold that against JS web apps.
flukus 10y ago
No, fastlite and a dozen other facebook apps have proven this incorrect. The official facebook app is slower, more bloated and battery draining than the mobile version ever was.
objclxt 10y ago
That article is four years old. Facebook's app architecture has changed significantly since then, as has the speed at which JS can execute.
matthewmacleod 10y ago
That article's no longer relevant, and in any case the existence of a bad implementation does not mean that a good implementation is not possible.
iainmerrick 10y ago
That seems irrelevant here -- Facebook is a web app that needs you to log in, and displays user-specific info in a very dynamic way. It's not the kind of thing that could ever be cached, whether by Amp or any other generic CDN.

manmal 10y ago

Two more reasons why this comparison is not on point: - On iOS, the FB App had to make do with a really slow JS engine at the time. IIRC Safari's JS performance was x5 compared to JS performance in apps' web views? This has more to do with how iOS treats apps, rather than how fast web pages in Safari are or were. - The FB apps were okayish at the time, given that they were web based. If you manage to get your web app to that level, your site is ok. Now add to that the performance gains and browser optimizations of the last 4 years and you get a system that flies. Native apps are still faster and respond quicker to touch, but you cannot hold that against JS web apps.

flukus 10y ago

No, fastlite and a dozen other facebook apps have proven this incorrect. The official facebook app is slower, more bloated and battery draining than the mobile version ever was.

objclxt 10y ago

That article is four years old. Facebook's app architecture has changed significantly since then, as has the speed at which JS can execute.

matthewmacleod 10y ago

That article's no longer relevant, and in any case the existence of a bad implementation does not mean that a good implementation is not possible.

iainmerrick 10y ago

That seems irrelevant here -- Facebook is a web app that needs you to log in, and displays user-specific info in a very dynamic way. It's not the kind of thing that could ever be cached, whether by Amp or any other generic CDN.