hims-clone VS data-client

Provides data fetching, but handles mutations better by using reactive paradigms.

You can still do SSR with normal client-side behavior if you use a React 18 SSR compatible library

https://resthooks.io/ is inspired by Backbone models

The control flow can be pretty straightforward once you have middleware integration; which you can see with async state management becoming as simple as a one-liner

You're probably better off using a tool that is built with code reuse in mind like https://resthooks.io

Probably the most important thing is to have a normalized data cache to save yourself tons of data-bug headaches. I use https://resthooks.io/ myself because it used inferred typing safety meaning I don't have to do any extra definitions or even use typescript and it will tell me what types I should expect. But you could also roll your own normalization using https://github.com/paularmstrong/normalizr with a state management tool like redux

If the data ever changes I prefer using Rest Hooks to ensure data integrity without having to manually write state updates for every usage

The advantage really comes when you using libraries like https://resthooks.io/ that enforce and infer strong types without effort from you. Then all you're doing is writing the types. Even cases like typos get removed due to typeahead helping you out. This is especially helpful in large projects with many developers as you'll likely be frequently working with types someone else defined and you don't know their members off the top of your head.

rest-hooks - Delightful data fetching for React.