Hacker News new | ask | show | jobs
by gpjanik 850 days ago
Serious question: what is the issue with these paritcular sizes? I know that features/look these websites have are definitely achievable with less JS at a higher engineering cost, but what's the problem with it? 10MB loads in two seconds on an okay-ish desktop connection (correct me if I'm wrong, but most of people don't deploy Vercel apps from their phone from a mountain range with 3G connection). The experience on the websites mentioned is smooth as it can only get; everything is super fast and nice. Every subsequent click is just instant action. That's how web should look like.

Is the problem here that they perform poorly on slower computers/connections? Is it even true? Is there an audience of developers who can't use Vercel or GitLab productively because of that? Any metrics to support that? IMHO optimizing against bundle size/JS sent over the network is one of the worst metrics for performance I could imagine.
3 comments
AHTERIX5000 850 days ago
But the experience is not smooth as it can get if you're running a bit older HW. Gmail and Slack are notorious for bad performance even when the content displayed is rather simple plain text. As a user I don't really get anything in return when developers decide to use complex JS solutions for simple use cases.
link
npteljes 850 days ago
I think it serves as a generic metric for bloat, because nobody really optimizes for size, thereby making it a good untainted metric. As the web gets bloaty and slow, the size of the websites grow as well, which also invites using size as a metric for bloat.

Smooth, fast, nice, these would be good to measure, but it's much harder. I like an interface response time metric, for example [0]. I always lament that interfaces are getting slow - I get that they are nicer, too, but god damn why am I waiting 1 second for anything, when my Pentium III with Win XP was near instant?

[0] https://www.nngroup.com/articles/response-times-3-important-...

link
alexey2020 850 days ago
came here to write a similar comment. Totally agree!

Focusing on a particular metric for the sake of the metric - what's the point?

Let's spend a couple months, refactor an app to generate less js just to look cool in the eyes of dev community?

link