[Hawkular-dev] [process] tickets and PRs

[Thomas Segismont]

I'm opposed to that as a rule.

If a PR has been under works for several days and impacts more than an 
handful of files, then I think it should be created ASAP so that the 
review can start early and be split in small commits.

I think you can't expect a good job from the reviewer when the diff says
"Showing 60 changed files with 3,593 additions and 2,530 deletions."
https://github.com/hawkular/hawkular-alerts/pull/50/files

Maybe I'm a slacker :)

Le 27/05/2015 13:11, Lucas Ponce a écrit :
> Let me explain myself better.
>
> Or in which situations I think squash is good:
>
> - I need to fix a HW-5555.
> - I checkout master and I create a new branch for HW-5555.
> - I start to work there:
> - My commits:
>    - Start working.
>    - Adding a couple of things.
>    - Javadoc typo.
>    - License agains...
>
> - I think this all 4 commits can be squashed without putting in danger anything.
>
> - We can send the PR to review without squash, and after that, squash to have a cleaner structure.
>
> Again, this can help IMHO.
>
> ----- Original Message -----
>> From: "John Mazzitelli" <mazz at redhat.com>
>> To: "Discussions around Hawkular development" <hawkular-dev at lists.jboss.org>
>> Sent: Wednesday, May 27, 2015 1:05:44 PM
>> Subject: Re: [Hawkular-dev] [process] tickets and PRs
>>
>> Uh...I recommend NO squash commits unless you are absolutely sure you know
>> what you are doing :)
>>
>> I can remember we screwed ourselves badly in the beginning of RHQ when we
>> used squashed commits wrong - the repository got all messed up.
>>
>> The only time I do a squash commit is if I immediately destroy the branch it
>> came from after the squash commit is performed, that way I know I won't
>> screw anything up.
>>
>> ----- Original Message -----
>>> I also propose to squash commits when it has sense.
>>>
>>> Project is growing and for debugging which "commit did that" grouping
>>> commits
>>> into a feature/fix/patch can help.
>>>
>>> Obviosly there are situations that are not possible, but it can be good to
>>> start thinking on that.
>>>
>>> Just my 2 cents.
>>> I can remember we screwed ourselves badly in the beginning of RHQ when we used squashed commits wrong - the repository got all messed up.
>
>>> ----- Original Message -----
>>>> From: "Thomas Segismont" <tsegismo at redhat.com>
>>>> To: hawkular-dev at lists.jboss.org
>>>> Sent: Wednesday, May 27, 2015 12:12:38 PM
>>>> Subject: Re: [Hawkular-dev] [process] tickets and PRs
>>>>
>>>> Le 27/05/2015 12:06, Thomas Segismont a écrit :
>>>>> as many of
>>>>> the commits to the branch have been sent as PR already.
>>>>
>>>> Well, we should have sent all commits as part of PR and related to a
>>>> ticket. That would have made the review process much easier than
>>>> reviewing PR 228
>>>> _______________________________________________
>>>> hawkular-dev mailing list
>>>> hawkular-dev at lists.jboss.org
>>>> https://lists.jboss.org/mailman/listinfo/hawkular-dev
>>>>
>>>
>>> _______________________________________________
>>> hawkular-dev mailing list
>>> hawkular-dev at lists.jboss.org
>>> https://lists.jboss.org/mailman/listinfo/hawkular-dev
>>>
>>
>> _______________________________________________
>> hawkular-dev mailing list
>> hawkular-dev at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/hawkular-dev
>>
>
> _______________________________________________
> hawkular-dev mailing list
> hawkular-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hawkular-dev
>