livedata-combineutil-java VS Clother

3.) the MutableLiveDatas in the ViewModel initialized by SavedStateHandle.getLiveData should contain the fromDate and the toDate, and with the combination of these LiveDatas you can evaluate what to show on the screen

well the third-party libraries help but internally it's just MediatorLiveData which does ship with LiveData

If you're using Google's defaults, then you use a ViewModel created by Hilt and getting a reference to a SavedStateHandle, your state would be created of multiple MutableLiveDatas (each holding 1 property) created via savedStateHandle.getLiveData(), then you use MediatorLiveData (you can use a library like this or this) to combine the changes of the mutable properties into a single immutable state object observed by the "UI" (read: the view controller, aka the Fragment) using the viewLifecycleOwner property of Fragments.

MutableLiveData (java combiner/kotlin combiner) == BehaviorRelay (kotlin combiner) == MutableSharedFlow (kotlin combiner) conceptually, so it's not important to me which one it is, as long as you can throw combiners at it.

Yep, that's why I have this kotlin combiner/java combiner for LiveData, but map/switchMap also rely on MediatorLiveData.

people exposing 11 public LiveData fields for 11 private MutableLiveData fields instead of using MediatorLiveData as combiners to reduce the number of exposed LiveData fields

While it is possible to combine these properties into a single emission using MediatorLiveData, if the MutableLiveData is exposed as MutableLiveData, then the fields can be directly two-way databound to XML, which is kinda the original goal and purpose of databinding.

If you need help with combiners you can also check https://github.com/Zhuinden/livedata-combineutil-java/

If you can't join them, you can combine LiveDatas and then map them to a common representation so that the adapter can treat it as a single list. And use itemViewType to determine which view to show per each item.

The main features are user authentication, paginated list of offers, search by text query or category with filters (size and location), creation of your own offer with up to 5 images, category, title, description, size and location, real-time p2p chat using Socket.IO with FCM push notifications support and caching of offers, chats and messages. You can find a more detailed description, short video demonstration and source code on Github: https://github.com/t3ddyss/Clother.