It seems that most paging libraries let you invalidate the dataset from the outside, i.e. pager.invalidate. Could anyone explain why thats the default in paging, compared to loading a
Flow<List<T>>
for each of its pages, that then gets combined into the final