getBreedsFromCache().conflate()

[hlnstepanova]

Hi Brady, could you please explain what is the purpose of conflate() in BreedModel getBreeds() in the last version of the project?
getBreedsFromCache().conflate().collect {
emit(it)
}

[brady-aiello]

Hi Lena, sorry for not responding sooner. Still getting used to GH discussions. The current version no longer uses conflate, but it could. According to the docs:

Conflates flow emissions via conflated channel and runs collector in a separate coroutine. The effect of this is that emitter is never suspended due to a slow collector, but collector always gets the most recent value emitted.

If we were liking / unliking breeds, and producing new updated versions of the DB faster than we could collect them, conflate() would allow us to always have the latest value, instead of having to wait for every intermediate value to be collected.

At the time, I was collecting inside the repository layer (BreedModel), which requires a coroutine scope to launch. The pro was that it could expose only a single flow to the ViewModel. The con was that it required passing it a coroutine scope. The current strategy is to return 2 flows from the repository layer, and flattenMerge() them into 1 in the ViewModel here.

[hlnstepanova]

I've noticed it changed, but was helpful to know why it used to be that way. Thanks for the explanation!