Improve JSON analytics data structure

Improve JSON analytics data structure according to e0101981 propositions below:

For me there are a few things what could be better:

1. Change applicationVersion => application, and have name & version inside it
   • Makes JSON structure more generic and we can later on utilize same on other apps than Qt creator?
2. Change qtVersion to qt, and put “version” inside it.
   • Maybe we want later on to have other properties of Qt as well. For example type: opensource, evaluation, commercial.
3. Somehow I would like to have generic & analytics properties separated. What I mean that underlying environment and actual analytics events would be separated, something like this (Did not really think it through but just idea):

 

    {

        "environment": {

            "application": {

                "name": "Qt Creator",

                "version": "4.5.2"

            }

            "platform": {

                "kernel": "4.15.13-1-default",

                "name": "openSUSE Tumbleweed",

                "os": "linux",

                "release": "20180326",

                "type": "opensuse"

            },

            "compiler": {

                "type": "GCC",

                "version": "7.3.1"

            }

            ...

        }

        "analytics": {

            "qmlDesignerUsageTime": {

                "value": 2173

            },

            "startCount": {

            "   value": 22

            },

        }

    }

   

4. I also wonder why to have qmlDesignerUsageTime as an object, if there is no other properties than value inside? Naming of property also efficiently prevent extending it. Why not have "qmlDesigner" property and inside it "usageTime", then you could easily add for example "startCount" inside it?