Skip to content

Latest commit

 

History

History
50 lines (28 loc) · 2.35 KB

performance.md

File metadata and controls

50 lines (28 loc) · 2.35 KB

🚄 Performance

Code Splitting

Code splitting is a technique of splitting production JavaScript into smaller files, thus allowing the application to be only partially downloaded. Any unused code will not be downloaded until it is required by the application.

Most of the time code splitting should be done on the routes level, but can also be used for other lazy loaded parts of application.

Do not code split everything as it might even worsen your application's performance.

Code Splitting Example Code

Component and state optimizations

  • Do not put everything in a single state. That might trigger unnecessary re-renders. Instead split the global state into multiple stores according to where it is being used.

  • Keep the state as close as possible to where it is being used. This will prevent re-rendering components that do not depend on the updated state.

  • If you have a piece of state that is initialized by an expensive computation, use the state initializer function instead of executing it directly because the expensive function will be run only once as it is supposed to. e.g:

// instead of this which would be executed on every re-render:
const [state, setState] = React.useState(myExpensiveFn());

// prefer this which is executed only once:
const [state, setState] = React.useState(() => myExpensiveFn());

How to write performant react code

  • Don't overuse useMemo and useCallback. Premature optimization can harm the overall performance of the app.

How to useMemo useCallback Before you use memo

Image optimizations

Consider lazy loading images that are not in the viewport.

Use modern image formats such as WEBP for faster image loading.

Use srcset to load the most optimal image for the clients screen size.

Web vitals

Since Google started taking web vitals in account when indexing websites, you should keep an eye on web vitals scores from Lighthouse and Pagespeed Insights.

Patterns

There is a lot of different patterns to optimize a production code. You have some references here Patterns.dev