Not sure how easy this is to explain: for a number of teams within my company, the definition of "done" is that the initial work (changeset) is "peer reviewed" (which comes after coding) but not functionally tested. All the statuses are correctly placed in the right columns and it works as expected for the most part. I understand that this is not what Scrum and Agile preach, we're basically saying that functional testing is not sprintable; we're working on making that a possibility but it is our current reality that will not change for the immediate future.
The problem: if a ticket passes peer review and thus is "done" within the Sprint but then fails in testing also within the Sprint and the developer starts working on the fix (which changes the status to something that doesn't count as "done") and doesn't finish before the end of the Sprint, at the end of the Sprint the ticket will not count as "done" although it meets our definition.
The question: is there any way to make this sticky as in once a ticket reaches a status that is mapped to the "done" column, if that status should change to something that maps to "In progress" the ticket is still considered "done" from a Sprint retrospective purpose? I understand that it would not move into the next Sprint, we have ways to keep track of our rework within our 2 Sprint release cycles.
Any help would be tremendously appreciated. Again, I understand that it's not ideal so please don't tell me to complete testing and rework within each Sprint, we'll get there, just not tomorrow. I just want to know if there's any way to use existing JIRA functionality to accomplish this. Thank you!
My advise, based upon experience, is to decouple testing from the dev. Let Dev use scrum and testing be independent. When dev is done then test can open bugs against the ‘feature’. This would be done by opening an issue and linking it to the done issue. You can even consider having a separate Kanban board for the teating tasks.
I agree with and understand the advice, but I have been unable to get buy-in on this approach. Both developers and business analysts (who do all our testing) have strongly pushed back on the overhead of creating new tickets for each follow-up changeset needed to address even a small failure to the original ticket. While not all tickets fail in testing, some do, some fail multiple times. So you could potentially end up with 3-4-5-6 tickets instead of 1. Any ideas on how to make this work in JIRA?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Well that is unfortunate on many fronts. That mindset is flawed for a number of reasons, e.g. you cant easily track the number of bugs against a feature/dev item. But let’s not turn this into a religious war. :-)
ultimately if the business isn’t going to embrace Agile then you should simply use Kanban and agile-like principals, e.g. short development cycles. In the current environment you won’t be able to use scrum reports to assess velocity, work done and undone.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
I understand that full well but again, Agile principles aside, can someone please help me with the question I asked in my post. Is there any JIRA functionality or "hack" that prevents a ticket from being considered not done if it reached a done status once?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
sorry i could not be of more help. good luck!!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.