incubator-stdcxx-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Martin Sebor <se...@roguewave.com>
Subject Re: JIRA issues: resolve or close or resolve then close ?
Date Wed, 21 Feb 2007 17:31:36 GMT
Farid Zaripov wrote:
>   Usually I used 'resolve' action.
>  
>   I thought that 'resolve issue' means 'do something to fix the issue and
> then close'
> and 'close issue' means 'do nothing and just close', but today I've looked
> into JIRA
> workflow
> (http://www.atlassian.com/software/jira/docs/v3.7.2/default_workflow.html
> <BLOCKED::http://www.atlassian.com/software/jira/docs/v3.7.2/default_workflo
> w.html> )
> and seen that 'closed' state is the final state of the issue.

That's correct.

>  
>   Shall I close all issues which are resolved by me?
>  
>   Do we have any rule on that matter?

Here's how I tend to work with issues:

  * I Open new issues and leave them Unassigned unless I know
    I will be working on it soon.
  * I Assign an Open issue to Myself when I know I will be
    working on it soon.
  * I Start Progress on an issue as soon as I start working
    on it. I Stop Progress on an issue when I go do something
    else and don't expect to get back to it for some time.
    (I confess I'm not very diligent about this though.)
  * I Resolve an issue after a fix has been committed and
    verified (i.e., test case passes). I leave the issue in
    Resolved state until a test for it has been added to
    the test suite.
  * I Close issues that have tests in the test suite which
    pass. I also Close issues (usually bypassing the Resolved
    state) that we can't easily add tests for (such as makefile
    bugs, etc.)

The one thing that's fuzzy in this process is the part about
adding the test case to the test suite. We have no regression
tests so I've been enhancing the conformance tests instead.
I don't think that's quite right. I think we should add
a regression test for each bug, even if the same functionality
is (or will be) also tested by a conformance test.

>  
>   Hmm, I was wrong: the 'closed' state is not final (it's just lowest at the
> state diagram)
> because an issue always can be reopened...

Right. It is the desired final state.

Martin


Mime
View raw message