Re: Jenkins Core pull-requests management

[Adrien Lecharpentier]

Hello everyone,

I created https://github.com/jenkinsci/jenkins/pull/6413 to move this 
subject forward. 

Le mercredi 23 mars 2022 à 10:39:39 UTC+1, Adrien Lecharpentier a écrit :

> For everyone, to give data about why I started this discussion:
>  - we have 66 opens pull-requests, 5 of them are marked as ready-for-merge
>  - 10 out of 61 pull-requests are labeled proposed-for-close
>  - of the 10 oldest open pull-requests, 7 are labels proposed-for-close
>  - the 17 oldest open pull-requests were all created in 2020.
>
> Some of those old ones have been reviewed, approved but yet, because they 
> are old, now have merge conflicts. As some authors change focus with time, 
> which is perfectly normal to me, at some point we don't get any activities, 
> leading the pull-requests to be stalled. 
> I think having a clear closing policy will also help us prevent this 
> situation.
>
> Le mer. 23 mars 2022 à 09:43, Adrien Lecharpentier <
> adrien.lec...@gmail.com> a écrit :
>
>> Alex, I agree that we have the list of labels but even for 
>> `ready-for-merge`, we haven't document there what we tend to say when we 
>> put the label: the pr should be merge within 24hr (with no negative 
>> feedback). It is important that we have a clear guideline, like 
>>
>> >  * If you do not get feedback after 3 days, feel free to ping 
>> `@jenkinsci/core-pr-reviewers` in the comments.
>> -- (
>> https://github.com/jenkinsci/jenkins/blob/e3f8bc052d6065d3fcf6430f4c7c029c1fd25e54/CONTRIBUTING.md?plain=1#L107
>> )
>>
>> I think it's important to be clear, for the community and all 
>> contribution that everyone opening a PR should get a review and comments. 
>> But also that, to limit the time lost by anyone, when a PR is no where 
>> close to be merge, there is no consensus on it's content, there are 
>> concerns about its quality / justifications, we should close it, in a 
>> timely matter. 
>>
>> Again, I think this is to show respect to all contributors, to show that 
>> contributions are not going into a void, that we care about them. And I 
>> feel like we are lacking clarity about it for the moment. 
>>
>> Le mar. 22 mars 2022 à 23:41, Olivier Lamy  a écrit :
>>
>>>
>>>
>>> On Wed, 23 Mar 2022 at 02:07, Adrien Lecharpentier <
>>> adrien.lec...@gmail.com> wrote:
>>>
 Hello everyone,

 I've spent some time lately on looking at the pull-requests on 
 jenkinsci/jenkins repository. For some old, inactive pull-requests I've 
 pinged the authors and for some, added the proposed-for-close label. 

 However, the label description nor any prior discussion on the 
 mailing-list are mentioning our policy about this (proposed-for-close) 
 label. And I'd like to offer one: I'd like, as for the ready-for-merge 
 label, to introduce a period of time, after which with no response from 
 the 
 author, we close the pull-request. I was thinking about 72 or 96hr.

>>>
>>> Definitely sounds good to clean up some stale/dead PRs but I find 96h a 
>>> bit too short. (people can be off for few days/weeks and 1 or 2 weeks for 2 
>>> yo old PR will not hurt more)
>>> I have implemented something similar in some plugins using stale action 
>>> see configuration here 
>>> https://github.com/jenkinsci/maven-plugin/blob/master/.github/workflows/stale.yml
>>>   
>>> (tool here: https://github.com/actions/stale)
>>> with this configuration PRs 365 days old are marked stale then after 30 
>>> days they are closed.
>>>
>>  
>> A PR can always be reopen, the discussions and work put in it is not 
>> lost. 96hr with no activities seems long enough. It's not 96hr after the 
>> last message / commit, but 96hr after the label *with a message* is applied 
>> to the PR. We could definitively say that for PR with no activities in the 
>> last month, or less if some review comment were not addressed, we put a 
>> message and the label on the PR.
>>  
>>
>>>  This might seem a bit harsh, but my idea is to try to keep the 
>>> pull-requests list healthy. And when we have no consensus on the PR or no 
>>> response from the authors, it's healthier to close the pull-request. The 
>>> work done is not lost, the PR can be reopen later when the author is more 
>>> available to attend to it. 
>>>

 Also, in case the authors respond, we can simply put the label stalled 
 (for others to take over the PR). We could also put the PR back into 
 draft, 
 but all members have enough permission to do that on others pull-requests, 
 but we could use the work-in-progress. Of course, after another period of 
 time, with no more activities, we should close the PR anyway.

 Also, there is no mention of those labels, what they means and how we 
 use it on our contribution guide [1]. Should we add a mention of those 
 label on it?

 -- Adrien

 [1]: https://github.com/jenkinsci/jenkins/blob/master/CONTRIBUTING.md

 -- 
 You received this message because you are 

Re: Jenkins Core pull-requests management

[Adrien Lecharpentier]

For everyone, to give data about why I started this discussion:
 - we have 66 opens pull-requests, 5 of them are marked as ready-for-merge
 - 10 out of 61 pull-requests are labeled proposed-for-close
 - of the 10 oldest open pull-requests, 7 are labels proposed-for-close
 - the 17 oldest open pull-requests were all created in 2020.

Some of those old ones have been reviewed, approved but yet, because they
are old, now have merge conflicts. As some authors change focus with time,
which is perfectly normal to me, at some point we don't get any activities,
leading the pull-requests to be stalled.
I think having a clear closing policy will also help us prevent this
situation.

Le mer. 23 mars 2022 à 09:43, Adrien Lecharpentier <
adrien.lecharpent...@gmail.com> a écrit :

> Alex, I agree that we have the list of labels but even for
> `ready-for-merge`, we haven't document there what we tend to say when we
> put the label: the pr should be merge within 24hr (with no negative
> feedback). It is important that we have a clear guideline, like
>
> >  * If you do not get feedback after 3 days, feel free to ping
> `@jenkinsci/core-pr-reviewers` in the comments.
> -- (
> https://github.com/jenkinsci/jenkins/blob/e3f8bc052d6065d3fcf6430f4c7c029c1fd25e54/CONTRIBUTING.md?plain=1#L107
> )
>
> I think it's important to be clear, for the community and all contribution
> that everyone opening a PR should get a review and comments. But also that,
> to limit the time lost by anyone, when a PR is no where close to be merge,
> there is no consensus on it's content, there are concerns about its quality
> / justifications, we should close it, in a timely matter.
>
> Again, I think this is to show respect to all contributors, to show that
> contributions are not going into a void, that we care about them. And I
> feel like we are lacking clarity about it for the moment.
>
> Le mar. 22 mars 2022 à 23:41, Olivier Lamy  a
> écrit :
>
>>
>>
>> On Wed, 23 Mar 2022 at 02:07, Adrien Lecharpentier <
>> adrien.lecharpent...@gmail.com> wrote:
>>
>>> Hello everyone,
>>>
>>> I've spent some time lately on looking at the pull-requests on
>>> jenkinsci/jenkins repository. For some old, inactive pull-requests I've
>>> pinged the authors and for some, added the proposed-for-close label.
>>>
>>> However, the label description nor any prior discussion on the
>>> mailing-list are mentioning our policy about this (proposed-for-close)
>>> label. And I'd like to offer one: I'd like, as for the ready-for-merge
>>> label, to introduce a period of time, after which with no response from the
>>> author, we close the pull-request. I was thinking about 72 or 96hr.
>>>
>>
>> Definitely sounds good to clean up some stale/dead PRs but I find 96h a
>> bit too short. (people can be off for few days/weeks and 1 or 2 weeks for 2
>> yo old PR will not hurt more)
>> I have implemented something similar in some plugins using stale action
>> see configuration here
>> https://github.com/jenkinsci/maven-plugin/blob/master/.github/workflows/stale.yml
>> (tool here: https://github.com/actions/stale)
>> with this configuration PRs 365 days old are marked stale then after 30
>> days they are closed.
>>
>
> A PR can always be reopen, the discussions and work put in it is not lost.
> 96hr with no activities seems long enough. It's not 96hr after the last
> message / commit, but 96hr after the label *with a message* is applied to
> the PR. We could definitively say that for PR with no activities in the
> last month, or less if some review comment were not addressed, we put a
> message and the label on the PR.
>
>
>>  This might seem a bit harsh, but my idea is to try to keep the
>> pull-requests list healthy. And when we have no consensus on the PR or no
>> response from the authors, it's healthier to close the pull-request. The
>> work done is not lost, the PR can be reopen later when the author is more
>> available to attend to it.
>>
>>>
>>> Also, in case the authors respond, we can simply put the label stalled
>>> (for others to take over the PR). We could also put the PR back into draft,
>>> but all members have enough permission to do that on others pull-requests,
>>> but we could use the work-in-progress. Of course, after another period of
>>> time, with no more activities, we should close the PR anyway.
>>>
>>> Also, there is no mention of those labels, what they means and how we
>>> use it on our contribution guide [1]. Should we add a mention of those
>>> label on it?
>>>
>>> -- Adrien
>>>
>>> [1]: https://github.com/jenkinsci/jenkins/blob/master/CONTRIBUTING.md
>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "Jenkins Developers" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to jenkinsci-dev+unsubscr...@googlegroups.com.
>>> To view this discussion on the web visit
>>> 

Re: Jenkins Core pull-requests management

[Adrien Lecharpentier]

Alex, I agree that we have the list of labels but even for
`ready-for-merge`, we haven't document there what we tend to say when we
put the label: the pr should be merge within 24hr (with no negative
feedback). It is important that we have a clear guideline, like

>  * If you do not get feedback after 3 days, feel free to ping
`@jenkinsci/core-pr-reviewers` in the comments.
-- (
https://github.com/jenkinsci/jenkins/blob/e3f8bc052d6065d3fcf6430f4c7c029c1fd25e54/CONTRIBUTING.md?plain=1#L107
)

I think it's important to be clear, for the community and all contribution
that everyone opening a PR should get a review and comments. But also that,
to limit the time lost by anyone, when a PR is no where close to be merge,
there is no consensus on it's content, there are concerns about its quality
/ justifications, we should close it, in a timely matter.

Again, I think this is to show respect to all contributors, to show that
contributions are not going into a void, that we care about them. And I
feel like we are lacking clarity about it for the moment.

Le mar. 22 mars 2022 à 23:41, Olivier Lamy  a écrit :

>
>
> On Wed, 23 Mar 2022 at 02:07, Adrien Lecharpentier <
> adrien.lecharpent...@gmail.com> wrote:
>
>> Hello everyone,
>>
>> I've spent some time lately on looking at the pull-requests on
>> jenkinsci/jenkins repository. For some old, inactive pull-requests I've
>> pinged the authors and for some, added the proposed-for-close label.
>>
>> However, the label description nor any prior discussion on the
>> mailing-list are mentioning our policy about this (proposed-for-close)
>> label. And I'd like to offer one: I'd like, as for the ready-for-merge
>> label, to introduce a period of time, after which with no response from the
>> author, we close the pull-request. I was thinking about 72 or 96hr.
>>
>
> Definitely sounds good to clean up some stale/dead PRs but I find 96h a
> bit too short. (people can be off for few days/weeks and 1 or 2 weeks for 2
> yo old PR will not hurt more)
> I have implemented something similar in some plugins using stale action
> see configuration here
> https://github.com/jenkinsci/maven-plugin/blob/master/.github/workflows/stale.yml
> (tool here: https://github.com/actions/stale)
> with this configuration PRs 365 days old are marked stale then after 30
> days they are closed.
>

A PR can always be reopen, the discussions and work put in it is not lost.
96hr with no activities seems long enough. It's not 96hr after the last
message / commit, but 96hr after the label *with a message* is applied to
the PR. We could definitively say that for PR with no activities in the
last month, or less if some review comment were not addressed, we put a
message and the label on the PR.


>  This might seem a bit harsh, but my idea is to try to keep the
> pull-requests list healthy. And when we have no consensus on the PR or no
> response from the authors, it's healthier to close the pull-request. The
> work done is not lost, the PR can be reopen later when the author is more
> available to attend to it.
>
>>
>> Also, in case the authors respond, we can simply put the label stalled
>> (for others to take over the PR). We could also put the PR back into draft,
>> but all members have enough permission to do that on others pull-requests,
>> but we could use the work-in-progress. Of course, after another period of
>> time, with no more activities, we should close the PR anyway.
>>
>> Also, there is no mention of those labels, what they means and how we use
>> it on our contribution guide [1]. Should we add a mention of those label on
>> it?
>>
>> -- Adrien
>>
>> [1]: https://github.com/jenkinsci/jenkins/blob/master/CONTRIBUTING.md
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "Jenkins Developers" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to jenkinsci-dev+unsubscr...@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/jenkinsci-dev/CAKwJSvwTnxuwe1WZzs3eSJBgq783fMm8hkQ_-%3DFHS1u0%2B7GUAw%40mail.gmail.com
>> 
>> .
>>
>
>
> --
> Olivier Lamy
> http://twitter.com/olamy | http://linkedin.com/in/olamy
>
> --
> You received this message because you are subscribed to the Google Groups
> "Jenkins Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to jenkinsci-dev+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/jenkinsci-dev/CAPoyBqTayEyLwyfXcaaE4bsSLOTj6YRPw5M%2BstBeiekq3F7TUA%40mail.gmail.com
> 
> .
>

-- 
You received this message because you are subscribed to the 

Re: Jenkins Core pull-requests management

[Olivier Lamy]

On Wed, 23 Mar 2022 at 02:07, Adrien Lecharpentier <
adrien.lecharpent...@gmail.com> wrote:

> Hello everyone,
>
> I've spent some time lately on looking at the pull-requests on
> jenkinsci/jenkins repository. For some old, inactive pull-requests I've
> pinged the authors and for some, added the proposed-for-close label.
>
> However, the label description nor any prior discussion on the
> mailing-list are mentioning our policy about this (proposed-for-close)
> label. And I'd like to offer one: I'd like, as for the ready-for-merge
> label, to introduce a period of time, after which with no response from the
> author, we close the pull-request. I was thinking about 72 or 96hr.
>

Definitely sounds good to clean up some stale/dead PRs but I find 96h a bit
too short. (people can be off for few days/weeks and 1 or 2 weeks for 2 yo
old PR will not hurt more)
I have implemented something similar in some plugins using stale action see
configuration here
https://github.com/jenkinsci/maven-plugin/blob/master/.github/workflows/stale.yml
(tool here: https://github.com/actions/stale)
with this configuration PRs 365 days old are marked stale then after 30
days they are closed.



>
> This might seem a bit harsh, but my idea is to try to keep the
> pull-requests list healthy. And when we have no consensus on the PR or no
> response from the authors, it's healthier to close the pull-request. The
> work done is not lost, the PR can be reopen later when the author is more
> available to attend to it.
>
> Also, in case the authors respond, we can simply put the label stalled
> (for others to take over the PR). We could also put the PR back into draft,
> but all members have enough permission to do that on others pull-requests,
> but we could use the work-in-progress. Of course, after another period of
> time, with no more activities, we should close the PR anyway.
>
> Also, there is no mention of those labels, what they means and how we use
> it on our contribution guide [1]. Should we add a mention of those label on
> it?
>
> -- Adrien
>
> [1]: https://github.com/jenkinsci/jenkins/blob/master/CONTRIBUTING.md
>
> --
> You received this message because you are subscribed to the Google Groups
> "Jenkins Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to jenkinsci-dev+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/jenkinsci-dev/CAKwJSvwTnxuwe1WZzs3eSJBgq783fMm8hkQ_-%3DFHS1u0%2B7GUAw%40mail.gmail.com
> 
> .
>


-- 
Olivier Lamy
http://twitter.com/olamy | http://linkedin.com/in/olamy

-- 
You received this message because you are subscribed to the Google Groups 
"Jenkins Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to jenkinsci-dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jenkinsci-dev/CAPoyBqTayEyLwyfXcaaE4bsSLOTj6YRPw5M%2BstBeiekq3F7TUA%40mail.gmail.com.

Re: Jenkins Core pull-requests management

[Alex]

Heyo,

that is an interesting proposal. Currently, there are quite a lot of stale 
pull requests, whether labeled with a stale-like label or not, which's 
state is unclear how to proceed with them. 
I think working with labels is easier, instead of converting the PRs into 
drafts, because a stale draft PR hardly differs from a draft that is work 
in progress. A label can show that state well without the need to read over 
the PR again.

> Also, there is no mention of those labels, what they means and how we use 
it on our contribution guide [1]. Should we add a mention of those label on 
it?
There's an overview of all labels available 
on https://github.com/jenkinsci/jenkins/labels. We could fill in missing 
descriptions and overhaul existing ones, if needed. The contribution guide 
could link to the former URL where you can browse PRs with specific labels 
right away.

~ Alex

On Tuesday, 22 March 2022 at 17:27:34 UTC+1 jef...@gmail.com wrote:

> This seems reasonable to me. After some period of time (months? years?) 
> the PR is unlikely to be merged, so why keep it open if it's truly stale?
>
>
> On Tue, Mar 22, 2022 at 9:07 AM Adrien Lecharpentier <
> adrien.lec...@gmail.com> wrote:
>
>> Hello everyone,
>>
>> I've spent some time lately on looking at the pull-requests on 
>> jenkinsci/jenkins repository. For some old, inactive pull-requests I've 
>> pinged the authors and for some, added the proposed-for-close label. 
>>
>> However, the label description nor any prior discussion on the 
>> mailing-list are mentioning our policy about this (proposed-for-close) 
>> label. And I'd like to offer one: I'd like, as for the ready-for-merge 
>> label, to introduce a period of time, after which with no response from the 
>> author, we close the pull-request. I was thinking about 72 or 96hr.
>>
>> This might seem a bit harsh, but my idea is to try to keep the 
>> pull-requests list healthy. And when we have no consensus on the PR or no 
>> response from the authors, it's healthier to close the pull-request. The 
>> work done is not lost, the PR can be reopen later when the author is more 
>> available to attend to it. 
>>
>> Also, in case the authors respond, we can simply put the label stalled 
>> (for others to take over the PR). We could also put the PR back into draft, 
>> but all members have enough permission to do that on others pull-requests, 
>> but we could use the work-in-progress. Of course, after another period of 
>> time, with no more activities, we should close the PR anyway.
>>
>> Also, there is no mention of those labels, what they means and how we use 
>> it on our contribution guide [1]. Should we add a mention of those label on 
>> it?
>>
>> -- Adrien
>>
>> [1]: https://github.com/jenkinsci/jenkins/blob/master/CONTRIBUTING.md
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "Jenkins Developers" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to jenkinsci-de...@googlegroups.com.
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/jenkinsci-dev/CAKwJSvwTnxuwe1WZzs3eSJBgq783fMm8hkQ_-%3DFHS1u0%2B7GUAw%40mail.gmail.com
>>  
>> 
>> .
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"Jenkins Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to jenkinsci-dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jenkinsci-dev/9dbeedc8-63a8-4c8e-b31f-252e8840d394n%40googlegroups.com.

Re: Jenkins Core pull-requests management

[Jeff]

This seems reasonable to me. After some period of time (months? years?) the
PR is unlikely to be merged, so why keep it open if it's truly stale?


On Tue, Mar 22, 2022 at 9:07 AM Adrien Lecharpentier <
adrien.lecharpent...@gmail.com> wrote:

> Hello everyone,
>
> I've spent some time lately on looking at the pull-requests on
> jenkinsci/jenkins repository. For some old, inactive pull-requests I've
> pinged the authors and for some, added the proposed-for-close label.
>
> However, the label description nor any prior discussion on the
> mailing-list are mentioning our policy about this (proposed-for-close)
> label. And I'd like to offer one: I'd like, as for the ready-for-merge
> label, to introduce a period of time, after which with no response from the
> author, we close the pull-request. I was thinking about 72 or 96hr.
>
> This might seem a bit harsh, but my idea is to try to keep the
> pull-requests list healthy. And when we have no consensus on the PR or no
> response from the authors, it's healthier to close the pull-request. The
> work done is not lost, the PR can be reopen later when the author is more
> available to attend to it.
>
> Also, in case the authors respond, we can simply put the label stalled
> (for others to take over the PR). We could also put the PR back into draft,
> but all members have enough permission to do that on others pull-requests,
> but we could use the work-in-progress. Of course, after another period of
> time, with no more activities, we should close the PR anyway.
>
> Also, there is no mention of those labels, what they means and how we use
> it on our contribution guide [1]. Should we add a mention of those label on
> it?
>
> -- Adrien
>
> [1]: https://github.com/jenkinsci/jenkins/blob/master/CONTRIBUTING.md
>
> --
> You received this message because you are subscribed to the Google Groups
> "Jenkins Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to jenkinsci-dev+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/jenkinsci-dev/CAKwJSvwTnxuwe1WZzs3eSJBgq783fMm8hkQ_-%3DFHS1u0%2B7GUAw%40mail.gmail.com
> 
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"Jenkins Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to jenkinsci-dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jenkinsci-dev/CADVhPTr%3DtRKmzoVfktCwGkUrLPKYPRV254wzWJEKMU0v0u-xwg%40mail.gmail.com.

Jenkins Core pull-requests management

[Adrien Lecharpentier]

Hello everyone,

I've spent some time lately on looking at the pull-requests on
jenkinsci/jenkins repository. For some old, inactive pull-requests I've
pinged the authors and for some, added the proposed-for-close label.

However, the label description nor any prior discussion on the mailing-list
are mentioning our policy about this (proposed-for-close) label. And I'd
like to offer one: I'd like, as for the ready-for-merge label, to introduce
a period of time, after which with no response from the author, we close
the pull-request. I was thinking about 72 or 96hr.

This might seem a bit harsh, but my idea is to try to keep the
pull-requests list healthy. And when we have no consensus on the PR or no
response from the authors, it's healthier to close the pull-request. The
work done is not lost, the PR can be reopen later when the author is more
available to attend to it.

Also, in case the authors respond, we can simply put the label stalled (for
others to take over the PR). We could also put the PR back into draft, but
all members have enough permission to do that on others pull-requests, but
we could use the work-in-progress. Of course, after another period of time,
with no more activities, we should close the PR anyway.

Also, there is no mention of those labels, what they means and how we use
it on our contribution guide [1]. Should we add a mention of those label on
it?

-- Adrien

[1]: https://github.com/jenkinsci/jenkins/blob/master/CONTRIBUTING.md

-- 
You received this message because you are subscribed to the Google Groups 
"Jenkins Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to jenkinsci-dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jenkinsci-dev/CAKwJSvwTnxuwe1WZzs3eSJBgq783fMm8hkQ_-%3DFHS1u0%2B7GUAw%40mail.gmail.com.