Saving Data on Android

compound sorting and filtering
. Additionally, in the Realtime 
Database, you cannot sort and filter the data in the same query, and when you query 
the data, the result is the whole subtree. Firestore allows all kinds of query chaining 
that NoSQL databases allow, and instead of querying entire collections or a 
document, you can query subcollections within a document. Furthermore, in the 
Realtime Database, you need to perform write operations in a single query; in the 
Firestore, you can collect all of your data and write it as a batch operation. This 
means that one large job is executed in small parts to improve efficiency.
Firestore has a lot of advanced features too. Both the Realtime Database and the 
Firestore offer offline support; however, the Realtime Database offers it only for the 
mobile clients, while the Firestore offers it for 
web apps
 as well.
In the Realtime Database, you need callbacks for transactions. In the Firestore, you 
don’t. Transactions are completed automatically until all of them are finished.
Scalability in the Realtime database isn’t that big of a problem, but when the data 
exceeds the limits you learned about in Chapter 15, “Usage and performance”, you 
needed to shred your data across multiple database instances. In the Firestore, you 
won’t need to do that regardless of how big your database will be — the