From f72f317cee081b752ee5100cb2599826264cd937 Mon Sep 17 00:00:00 2001
From: Vishesh Handa <me@vhanda.in>
Date: Thu, 5 Nov 2020 14:48:39 +0100
Subject: [PATCH] Custom analytics experiment

---
 lib/analytics_exp.dart | 67 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 67 insertions(+)
 create mode 100644 lib/analytics_exp.dart

diff --git a/lib/analytics_exp.dart b/lib/analytics_exp.dart
new file mode 100644
index 00000000..cda24dd0
--- /dev/null
+++ b/lib/analytics_exp.dart
@@ -0,0 +1,67 @@
+// - https://support.google.com/firebase/answer/7029846?hl=en
+// - https://support.google.com/firebase/answer/6317485?hl=en
+
+class Event {
+  DateTime date;
+  String name;
+  Map<String, dynamic> params;
+
+  String userId;
+  String psuedoId;
+  Map<String, dynamic> userProperties;
+
+  // Unique session identifier (based on the timestamp of the session_start
+  // event) associated with each event that occurs within a session
+  String sessionID;
+
+  String platform;
+  DateTime userFirstTouchTimestamp;
+}
+
+class Device {
+  String category; // mobile
+  String mobileBrandName;
+  String mobileModelName;
+  String mobileOsHardwareModel;
+  String operatingSystem;
+  String operatingSystemVersion;
+  String vendorId;
+  String language;
+  bool isLimitedAdTracking;
+  int timeZoneOffsetSeconds;
+}
+
+class Geo {
+  String continent;
+  String country;
+  String region;
+  String city;
+  String subContinent;
+  String metro;
+}
+
+class AppInfo {
+  String id;
+  String version;
+  String firebaseAppId;
+  String installSource;
+}
+
+//
+// * Create a postgres table with all this data
+//   -> Maybe use clickhouse instead?
+// * Convert all of this to the json representation we want
+// * Write a converter for the firebase data to this data format
+// * Insert all of this data into our clickhouse
+
+// * Create the dashboard - huge task
+// For the dashboard - maybe you could hire someone?
+// - Or I could just use Grafana
+// - Either way - it doesn't make sense to invest time in it
+
+// TODO -
+// - Create gj versions of the other events
+// - Store the events in a local db - use hive
+// - Populate the other data
+// - Post them to an endpoint which collects them
+// -