How to Design a Good Realtime Data Model in your own Collabrative App

Use Case Requirement of my Mobile App:

Let us consider the following use case of my mobile app to design a good realtime data model.

Use case: Reconcile chat messages in multiple devices. I need to show up history data generated by one device in another device in real time when user login in two different devices using same account details.

To implement this we need a central storage server and a synchronization service.

Solution:

Google Drive can be used as a file storage server and also I need to look for a javascript library or we should implement Operational Transformations (OT) that would allow us to synchronize history data in realtime between multiple devices. And Operational Transformations seems to fit my need for realtime sync. Google Drive Realtime API provides synchronization service for files in Google Drive via the use of Operational Transforms. This API is a JavaScript library hosted by google that provides collaborative objects, events, and methods for creating collaborative applciations.

The user has to authorize App access to his Google Drive. The following details explains how i did realtime data synchronization for my phonegap application and similarly you can do for any applicaitions (Web applications, Mobile applications and PhoneGap applications).

As I explained Google Drive Realtime API takes care of everything. Only thing that you have to do is to design a good Shared/Realtime/Collaborative data model and focus on how you can build your app. I have designed the below shared data model for my app.

Designing a Good Realtime Data Model:

The following is the proposed data model for my mobile application to implement realtime sync with Google Drive.

The Realtime/Shared/Collaborative Map entry contains a key and value pair. The Map's key must be a String. My mobile app/ chat application uses Cloud reference id as key which is generated using the combination of user's jabber id and timestamp of the chat session. The values can contain realtime collaborative objects. Here the value of the map would be collaborative list that contains the messages for the particular chat session.

Changes to the collaborative map will automatically be synced with ther server and other collaborators. To listen for changes we need to add EventListeners for the gapi.drive.realtime.EventType.VALUE_CHANGED event type.

Changes to the list will automatically be synced with the server and other collaborators. To listen for changes, add EventListeners for the following event types:

gapi.drive.realtime.EventType.VALUES_ADDED

gapi.drive.realtime.EventType.VALUES_REMOVED

gapi.drive.realtime.EventType.VALUES_SET

First entry in the collaborative list should always contain meta data about my app chat history session which includes cloud reference id, reciever jabber id, isActive flag , initiator device id etc. The remaining entries in the list are actual IM chat messges.

Web SQL Realtime Sync With Google Drive:

The above Realtime/Shared data model is designed based on the local database (WebSQL Storage)/ Data Models of my mobile application. The data model for chat history of my mobile application has the following two entities/tables: 

1. History table (For storing Chat Sessions)
2. IM table (For storing Messages)

Each chat session contains collection of messages chat by users. The session can be two party chat session or multiparty chat session. Whenever a collaborative data model is modified. The API then silently sends a representation of the change to the server (called a "mutation") so that the change can be recorded in the realtime file and communicated to other collaborators.

So we have discussed in detail how to design collaborative/realtime data model by understanding your local storage/ local data models. And also discussed how to do realtime data synchronization between two or more collaborators.