JIRA + BigPicture (Dev + QA workflow)

[pawelbah]

Hi,
I have this problem and I don't really know how to solve it.

We are using JIRA+BigPicture for organization and scheduling.
We have the structure organized in such a way that JIRA projects aggregate the work of teams (e.g. projects: android, ios, server, bigdata...). In projects, tasks are added for teams and labeled accordingly. Then, proper projects are created in BigPicture that pull specific tasks from individual JIRA projects (e.g., project prog-1 pulls all tasks with label prog-1).

QA-we have them organized in such a way that they are part of the workflow. The developer finishes the job and by changing the status passes it to the tester within the same project and workflow and here is the crux of the problem.

The problem is that the task is one and it is done by two people (programmer and tester). The estimate includes the work of the programmer and the tester. Problems:
- in BigPicture we see the workload of the person who is currently assigned (so the programmer is loaded with the time scheduled for the tester)
- as it is one assignment, so the tests are in the same block. You cannot, for example, start scheduling them for a few days later.

How do you solve this in your projects?

[Tomek Lewandowski]

Hi @pawelbah 

you might be interested in this film - an interesting combination of a test management tool and a product management tool

https://www.youtube.com/watch?v=ZZ5IJ8kTUTM 

cheers, 

Tomek

[Anna-BigPicture]

Hi @pawelbah

The first advice that comes to my mind is to divide a task into 2 subtasks or split a task in two and link them ASAP-mode link. This will allow you to assign issues to an accurate team and BigPicture will display the correct workload. 

[pawelbah]

The problem is that now you will have to add 2 artificial subtasks each time:
1. for the programmer
2. for the tester

Additionally, you have to remove the test phase from the workflow and create 2 separate ones:
1. for programmer with code review path
2. for tester (simple todo, in progress, done)

Another problem I notice is undoing the task by the tester. With such a setup, he would have to undo his subtask and undo the programmer's task.

It's a bit tricky :/