11:17 a.m.
We seem to have a few different understandings of Closed versus Resolved
as a status. We should come to a consensus.
According to Atlassian...
* Resolved : A resolution has been taken, and it is awaiting
verification by reporter. From here issues are either reopened, or are
closed.
* Closed : The issue is considered finished, the resolution is correct.
Issues which are not closed can be reopened.
For the most part I agree with these definitions.
The first thing to notice is that these definitions say (imply) that a
Closed issue cannot be reopened. Today, we do allow that. Do we really
want to not allow closed issues to be reopened? Partially this relies
on a few other decisions...
Should all issues that are part of a release be Closed? After an issue
has been part of a release, the normal process to do additional work on
the issue is to open a new issue. Mainly that is to keep release notes
straight. Unfortunately, as far as I can tell this would have to just
be a manual release step to transition all Resolved issues to Closed.
But I think from a release tracking perspective this is the most correct
option.
Comments. Today we allow comments on Closed issues. Unfortunately we
have to reopen the issue to then delete comments. So if we no longer
physically allow Closed issues to be reopened, then IMO we should
certainly not allow comments on Closed issues.
Which brings us to... do we really want to not allow Closed issues to be
reopened? I mean actually remove the transition? Or do we just want to
say that as a standard procedure we do not reopen Closed issues?
--
steve(a)hibernate.org
http://hibernate.org
11:40 a.m.
On Dec 20, 2011, at 11:17 AM, Steve Ebersole wrote:
Which brings us to... do we really want to not allow Closed issues to
be
reopened? I mean actually remove the transition? Or do we just want to
say that as a standard procedure we do not reopen Closed issues?
I think we should not allow them to be reopened. Seems like any type of comment that
really deserves attention also deserves a new issue linked to the closed issue.
JPAV
8:25 a.m.
On 12/20/2011 11:40 AM, John Verhaeg wrote:
On Dec 20, 2011, at 11:17 AM, Steve Ebersole wrote:
> Which brings us to... do we really want to not allow Closed issues to be
> reopened? I mean actually remove the transition? Or do we just want to
> say that as a standard procedure we do not reopen Closed issues?
I think we should not allow them to be reopened. Seems like any type of comment that
really deserves attention also deserves a new issue linked to the closed issue.
The question here is *who*? Who don't you think should be allowed to
reopen issues? The user? They cannot already. But I think totally
disallowing the development team from reopening closed issues is asking
for trouble.
--
steve(a)hibernate.org
http://hibernate.org
1:18 p.m.
Hi
On Dec 20, 2011, at 18:17, Steve Ebersole <steve(a)hibernate.org> wrote:
According to Atlassian...
* Resolved : A resolution has been taken, and it is awaiting
verification by reporter. From here issues are either reopened, or are
closed.
* Closed : The issue is considered finished, the resolution is correct.
Issues which are not closed can be reopened.
I agree with this as well. In HV and SEARCH we close the resolved issues after a release.
While they are resolved they could be reopened and closed issues cannot be reopened again.
A new issue should be created in this case.
Closing the issues is a manual process, but I don't see a problem with that. Compared
to other time wasters I don't mind the 2 minutes this process takes.
Comments. Today we allow comments on Closed issues. Unfortunately
we
have to reopen the issue to then delete comments. So if we no longer
physically allow Closed issues to be reopened, then IMO we should
certainly not allow comments on Closed issues.
What is the problem with comments on closed issues? Some comments might even be useful
(aka "had the same problem, fixed now") However, if someone wants to start a new
discussion - tough luck. If the comment is legit we tell them to create a new issue and
link to it.
Which brings us to... do we really want to not allow Closed issues to
be
reopened? I mean actually remove the transition?
That's a possibility. Can you do that in Jira?
--hardy
2:55 p.m.
I agree as well on interpreting "Closed" as "forever"; AFAIR new
issues can be linked to the older closed ones so that while a comment
might not be allowed, it's still possible to point to "follow ups" in
case of need.
I'm not sure however about enforcing this, but I'm not against it. I'd
prefer if committers had the option to reopen, it's never nice when
tools prevent you to fix eventual mistakes.
@Hardy yes we can customize JIRA to a great extent, especially in this area.
On the release steps: I'm not sure re Maven plugins, but JIRA has a
REST API as well for automating transitions. I agree with Hardy that
I'm not concerned about such a manual step, but in case we want in
future this could be automated as I think releases should be more and
more automated, but we can proceed by small steps; I'd expect this to
be easier with gradle, and maybe Maven plugins exists.
There definitely is automation in Jenkins which we could use for
release automation anyway:
https://wiki.jenkins-ci.org/display/JENKINS/Jira+Issue+Updater+Plugin
Cheers,
Sanne
On 20 December 2011 20:18, Hardy Ferentschik <hardy(a)hibernate.org> wrote:
Hi
On Dec 20, 2011, at 18:17, Steve Ebersole <steve(a)hibernate.org> wrote:
> According to Atlassian...
> * Resolved : A resolution has been taken, and it is awaiting
> verification by reporter. From here issues are either reopened, or are
> closed.
> * Closed : The issue is considered finished, the resolution is correct.
> Issues which are not closed can be reopened.
I agree with this as well. In HV and SEARCH we close the resolved issues after a release.
While they are resolved they could be reopened and closed issues cannot be reopened again.
A new issue should be created in this case.
Closing the issues is a manual process, but I don't see a problem with that. Compared
to other time wasters I don't mind the 2 minutes this process takes.
> Comments. Today we allow comments on Closed issues. Unfortunately we
> have to reopen the issue to then delete comments. So if we no longer
> physically allow Closed issues to be reopened, then IMO we should
> certainly not allow comments on Closed issues.
What is the problem with comments on closed issues? Some comments might even be useful
(aka "had the same problem, fixed now") However, if someone wants to start a new
discussion - tough luck. If the comment is legit we tell them to create a new issue and
link to it.
> Which brings us to... do we really want to not allow Closed issues to be
> reopened? I mean actually remove the transition?
That's a possibility. Can you do that in Jira?
--hardy
_______________________________________________
hibernate-dev mailing list
hibernate-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hibernate-dev
1:18 a.m.
Comments below...
----- Original Message -----
From: "Sanne Grinovero" <sanne(a)hibernate.org>
To: "Hardy Ferentschik" <hardy(a)hibernate.org>
Cc: hibernate-dev(a)lists.jboss.org
Sent: Tuesday, December 20, 2011 12:55:38 PM
Subject: Re: [hibernate-dev] JIRA issue status - Closed v. Resolved
I agree as well on interpreting "Closed" as "forever"; AFAIR new
issues can be linked to the older closed ones so that while a comment
might not be allowed, it's still possible to point to "follow ups" in
case of need.
I agree that "Closed" issues should not be reopened. Links should still be
possible for identifying duplicates, issues fixed/caused by a closed issued, etc.
I'm not sure however about enforcing this, but I'm not against it.
I'd
prefer if committers had the option to reopen, it's never nice when
tools prevent you to fix eventual mistakes.
I think it would be better to make it a convention that closed issues not be re-opened,
but to still allow committers to reopen it if an issue was closed by mistake.
After thinking about it more, I thing Hardy is right about allowing comments on closed
issues. I'd forgotten that we can actually filter on "Last commented by a
user". Aside from comments about problems with a fix, it's good to hear that the
fix was verified by a user and it's always nice to get a "thank you"
comment. :)
@Hardy yes we can customize JIRA to a great extent, especially in
this area.
I think it's a good idea to automatically close resolved issues when the fix is
released.
On the release steps: I'm not sure re Maven plugins, but JIRA has
a
REST API as well for automating transitions. I agree with Hardy that
I'm not concerned about such a manual step, but in case we want in
future this could be automated as I think releases should be more and
more automated, but we can proceed by small steps; I'd expect this to
be easier with gradle, and maybe Maven plugins exists.
There definitely is automation in Jenkins which we could use for
release automation anyway:
https://wiki.jenkins-ci.org/display/JENKINS/Jira+Issue+Updater+Plugin
Cheers,
Sanne
On 20 December 2011 20:18, Hardy Ferentschik <hardy(a)hibernate.org>
wrote:
> Hi
>
> On Dec 20, 2011, at 18:17, Steve Ebersole <steve(a)hibernate.org>
> wrote:
>
>> According to Atlassian...
>> * Resolved : A resolution has been taken, and it is awaiting
>> verification by reporter. From here issues are either reopened, or
>> are
>> closed.
>> * Closed : The issue is considered finished, the resolution is
>> correct.
>> Issues which are not closed can be reopened.
>
> I agree with this as well. In HV and SEARCH we close the resolved
> issues after a release. While they are resolved they could be
> reopened and closed issues cannot be reopened again. A new issue
> should be created in this case.
>
> Closing the issues is a manual process, but I don't see a problem
> with that. Compared to other time wasters I don't mind the 2
> minutes this process takes.
>
>> Comments. Today we allow comments on Closed issues.
>> Unfortunately we
>> have to reopen the issue to then delete comments. So if we no
>> longer
>> physically allow Closed issues to be reopened, then IMO we should
>> certainly not allow comments on Closed issues.
>
> What is the problem with comments on closed issues? Some comments
> might even be useful (aka "had the same problem, fixed now")
> However, if someone wants to start a new discussion - tough luck.
> If the comment is legit we tell them to create a new issue and
> link to it.
>
>> Which brings us to... do we really want to not allow Closed issues
>> to be
>> reopened? I mean actually remove the transition?
>
> That's a possibility. Can you do that in Jira?
>
> --hardy
>
> _______________________________________________
> hibernate-dev mailing list
> hibernate-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hibernate-dev
_______________________________________________
hibernate-dev mailing list
hibernate-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hibernate-dev
8:51 a.m.
On 12/20/2011 02:55 PM, Sanne Grinovero wrote:
I agree as well on interpreting "Closed" as
"forever"; AFAIR new
issues can be linked to the older closed ones so that while a comment
might not be allowed, it's still possible to point to "follow ups" in
case of need.
I'm not sure however about enforcing this, but I'm not against it. I'd
prefer if committers had the option to reopen, it's never nice when
tools prevent you to fix eventual mistakes.
Yeah I think this is the conclusion I am coming (back) to as well.
Which luckily is exactly how we have jira set up currently :)
I think the big change here (for core at least as it sounds like the
other projects do this already) is to consider released issues closed,
to the point of moving issues from a Resolved status to a Closed status
as part of the release.
By convention I still think an issue I report and then solve is Closed
rather than Resolved, but eh, its an opinion.
So overall:
* Close all issues on release
* Closed issues are, by *convention*, not reopened
* Allow dev team to reopen issues that are closed, but with the
understanding that this is there just to edit the issue details to fix
mistakes.
On the release steps: I'm not sure re Maven plugins, but JIRA has
a
REST API as well for automating transitions. I agree with Hardy that
I'm not concerned about such a manual step, but in case we want in
future this could be automated as I think releases should be more and
more automated, but we can proceed by small steps; I'd expect this to
be easier with gradle, and maybe Maven plugins exists.
There definitely is automation in Jenkins which we could use for
release automation anyway:
https://wiki.jenkins-ci.org/display/JENKINS/Jira+Issue+Updater+Plugin
Like I said in Hardy's response, my concern is not the time it takes to
do this, but more me actually remembering to do it :)
I will look at automating this via gradle at some point. Though, TBH,
setting up a jenkins job to do the release sounds fantastic. My concern
is the infinispan tests though. For each release lately I have had to
run the build multiple times until those happen to pass.
--
steve(a)hibernate.org
http://hibernate.org
8:30 a.m.
On 12/20/2011 01:18 PM, Hardy Ferentschik wrote:
I agree with this as well. In HV and SEARCH we close the resolved
issues after a release. While they are resolved they could be reopened and closed issues
cannot be reopened again. A new issue should be created in this case.
Closing the issues is a manual process, but I don't see a problem with that. Compared
to other time wasters I don't mind the 2 minutes this process takes.
The concern
here is not the time it takes; the concern is the actually
remembering to do it. I agree its not a major deal, but it is something
to consider.
What is the problem with comments on closed issues? Some comments
might even be useful (aka "had the same problem, fixed now") However, if someone
wants to start a new discussion - tough luck. If the comment is legit we tell them to
create a new issue and link to it.
In the past we had problems with spam comments showing up on closed
issues. It was just a pain in the ass to remove them because you had to
reopen the issue, remove the comment and then (re)close it.
In general I definitely agree that allowing comments on closed issues is
positive. In fact I would argue that it is necessary. In which case I
would say that we certainly want to still allow developers to reopen
closed issues.
--
steve(a)hibernate.org
http://hibernate.org
8:46 a.m.
On Dec 21, 2011, at 3:30 PM, Steve Ebersole wrote:
> What is the problem with comments on closed issues? Some comments
might even be useful (aka "had the same problem, fixed now") However, if someone
wants to start a new discussion - tough luck. If the comment is legit we tell them to
create a new issue and link to it.
In the past we had problems with spam comments showing up on closed issues. It was just
a pain in the ass to remove them because you had to reopen the issue, remove the comment
and then (re)close it.
I see. That's the first reason I hear for considering really forbidding comments on
closed issues. IMO it is really a Jira bug that as admin you cannot remove a comment
unless you reopen the issue.
Luckily we have not seen so much spamming on HV or SEARCH (I think)
Speaking of Jira and closed issues etc. I did a quick search today w/o any JIRA key. I was
expecting a search in the active projects (aka HHH, SEARCH, HV), but I got hits in
projects like TODO and HB. Is there a way to
exclude them from a default search?
In general I definitely agree that allowing comments on closed issues
is positive. In fact I would argue that it is necessary. In which case I would say that
we certainly want to still allow developers to reopen closed issues.
+1
--Hardy
4773
days inactive
4774
days old
8 comments
5 participants
participants (5)
-
Gail Badner -
Hardy Ferentschik -
John Verhaeg -
Sanne Grinovero -
Steve Ebersole