Closed glynnbird closed 6 years ago
Just to clarify the issue, when I'm on triage duty, there are 5 I might need to do, and 3 buttons with which to accomplish them:
The user doesn't know how the data is being captured and used. For example, they don't know that clicking "Answer" makes the system assume it's them who answered it, just that it marks the question as answered so we can all ignore it moving forward. They also don't know that the system is tracking how many questions people have answered on SO and getting that wrong if they don't follow the correct procedures for assignment while marking something answered.
As I understand it, there are two ways we could pull stats on how many questions each person on our team has answered:
For the record, the structure of documents in our database look like this: https://github.com/ibm-watson-data-lab/sodashboard/wiki/Repository-document-structures
assigned: true
and values set in assigned_by
, assigned_by_name
, assigned_at
. Code here: https://github.com/ibm-watson-data-lab/sodashboard/blob/master/public/js/sodashboard.js#L487-L514rejected:true
and values set in rejected_by
, rejected_by_name
, rejected_at
. Code here: https://github.com/ibm-watson-data-lab/sodashboard/blob/master/public/js/sodashboard.js#L515-L536answered: true
and values set for answered_by
, answered_by_name
, answered_at
. Code here: https://github.com/ibm-watson-data-lab/sodashboard/blob/master/public/js/sodashboard.js#L537-L557I've updated the wiki instructions for SO duty to clarify that when a question is answered, we just just click the answered button; we'll pull data about who answered what from the API.
If a user is triaging a SO ticket and finds that the question has already been answered by a non IBMer, they may be tempted to click the "Answered" button because it has been "Answered".
The intention of the "Answered" button is to record the fact that the triaging user answered the question. This is causing the stats to get skewed.
The solution is to rethink the UI a bit to make it unambiguous.