There was an interesting Twitter conversation recently where Paul Henschel showed a cool demo doing some 3D rendering of a crazy amount of data in React. He was using concurrent mode, an experimental new feature made to improve React Performance.
On its own, the demo was extremely impressive. But then Rich Harris (of whom I'm a big fan) did a similar thing in SvelteJS. Only he didn't do any tuning using experimental features and did it in dev mode (forgot to put it in optimized prod mode) and that demo seemed to be even more performant than the React demo.
A large amount of conversation ensued, lots of it technical about performance and such. I enjoyed watching the conversation as I'm a big fan of both React and Svelte.
Now there are a few things one could take away from this conversation.
You could look at this as evidence of which is the fastest framework.
But is that really the right lesson to take from this?
Take a step back and look at the bigger picture. Is performance the primary thing that matters in a framework (or most other tools for that matter)? Due to the limitations of humans, client machines/browsers, and internet bandwidth, it's easy to have far more performance than we can benefit from. Like driving a formula one race car in rush hour traffic. And it doesn't matter how fast you're going if you're heading down the wrong road.
I believe there's a better lesson to learn: and that is that frameworks are amazing, and their authors and contributors are doing cool stuff we're all benefiting from, but there's no single objective measure that makes a framework superior to its competitors. Taking a cue from what Andrew Clark said in that conversation, I choose to avoid the performance war hype.
Signup for my newsletter at here.