[Meta] Statistics
(This ticket is here to keep track of all export related issue)
Description
There are lot of data we are eager to collect, regarding both the hardware and the usage made by our users. This could prove very useful to make some design decisions. But, we should avoid spreading the myth that "collected data is anonymous so we your privacy is respected".
Warning
Privacy based on anonymity is a lie big (and smaller) tech companies tell people to gather any possible data and actually spy on them. We should keep in mind that the more data we collect, the more we are able to actually identify, or make profiles of the folks Paperwork is build for. Therefore, I wonder how we could avoid fingerprinting our users?
I mean, we certainly want to know, generally speaking, what is the more commonly used hardware by Paperwork users (CPU, RAM, storage, scanners, etc.). But we probably don't need to know for each single user what is the specific hardware the person owns, except perhaps for debugging purpose. The same goes for stats about documents, labels, etc.
Issues
UI:
-
#720 (closed) [statistics] indicates what is sent
Collect:
-
#705 [Statistics] collect the locale -
#710 [Statistics] Collect label guessing accuracy -
#717 [statistics] add CPU frequency + total amount of ram -
#768 [statistics] get screen resolution(s) -
#802 [statistics] document statistics
Analysis:
-
#789 (closed) [statistics] historic (watch progression over time) -
#768 [Statistics] Group Paperwork versions