Hacker News new | ask | show | jobs
by benmccann 1075 days ago
INP feels like a pretty problematic way to compare sites because INP is going to be way lower on a site that doesn't do client-side rendering eventhough client-side rendering makes interaction with a site faster!
3 comments

> client-side rendering makes interaction with a site faster!

I am going to have to disagree. Final HTML from the server is just that. Its final. The client displays it and its done. No XHR, no web sockets, no JS eval. It's done. You can immediately use the webpage and the webserver doesn't care who you are anymore. With SPA, this is the best case. You maybe even start with SSR from the server and try to incrementally move from there. Regardless, the added complexity of SSR->SPA and other various hybrid schemes can quickly eat into your technical bullshit budget and before you know it that ancient forms app feels like lightning compared to the mess you proposed.

Reaching for SPA because you think this will make the site "faster" is pretty hilarious to me. I've never once seen a non-trivial (i.e. requires server-side state throughout) SPA that felt better to me than SSR.

> I've never once seen a non-trivial (i.e. requires server-side state throughout) SPA that felt better to me than SSR.

What about gmail? That has all the state server side. How impressive would it be if all rendering was done server side?

People don’t care whether your site is server or client rendered… they care about fast interactions
I completely disagree. Client side has the potential to be very fast, even faster. However, most people are more interested in writing a complex, Turing complete, type system under their client than making fast, easy to use applications.