Re:“[Twitter] is doing 1000 poorly batched RPCs just to rend...
by csande17 on Sunday November 13, 2022 @07:38PM
(#33588927)
Pipelining and multiplexing can certainly help make multiple requests faster, but they aren't a magic bullet. In particular, if your frontend makes a lot of requests sequentially (which is pretty common in badly designed frontends -- a component makes a request, waits for the response before rendering children, a child makes a request, etc...), that's still going to perform really badly in high-latency environments.
Of course, it's equally possible to mess up your server-side logic and make it really slow. And a lot of teams have really low performance expectations, or expect all of their customers to have 5G internet connections and live next to a data center, and either of those can buy you time to make requests.
Re:“[Twitter] is doing 1000 poorly batched RPCs just to rend...
by lost_tourist on Monday November 14, 2022 @02:24PM
(#33599490)
I'm honestly curious, what is opportunistic model hydration? Short explanation would be great, I don't need all the gory details.
We’re currently in the process of ripping out a lot of client side opportunistic model hydration in favor of just making multiple requests because it’s faster.