LiveData or StateFlow? I don't understand why people use StateFlow. And there's flow for Repository....

# flow

[chanjungskim]

LiveData or StateFlow? I don't understand why people use StateFlow. And there's flow for Repository. I am very confused. Anybody can make me understand them...?

[CLOVIS]

You can't use LiveData server-side, in a JVM Desktop application, in your JS website, or in iOS.

[chanjungskim]

So, LiveData is just for android and StateFlow for other platforms?

[chanjungskim]

That's interesting that server side also uses StateFlow. Does it need to observe data?

[CLOVIS]

StateFlow is not for "other platforms". It's for all platforms, you only need to learn one thing and use it everywhere.

[CLOVIS]

Does the server need to observe data?

Depends on your server. I use flows to cache frequent DB requests in-memory → https://gitlab.com/opensavvy/pedestal/-/blob/main/cache/src/commonMain/kotlin/Cache.kt

[CLOVIS]

Since flows are multiplatform, I can use the same Cache implementation server-side to cache the DB and in the JS code to cache network requests. There's no need anymore to reinvent the wheel on every app.

[chanjungskim]

That's a good point. Thanks for the explanation. 🙂

[fuad]

Also Google mentions it's not necessary to mix LiveData with Flows when using Compose etc...they specifically say using LiveData isn't necessary and just an extra thing. Consuming a Flow from a Composable seems a bit cleaner to me than LiveData as well, and it goes along with the Compose MVI architecture

[chanjungskim]

For me, it's very confusing. It'd be nice if there are commonly used template or something.. I feel like I need to use flow and StateFlow but it's hard to apply..

[fuad]

There are! check out the Google sample projects