How about separate commit action between initial committers and new
committers?
For initial committers, let them use `git rebase -i` to management their
commits; for new committers we just use `squash and merge`.
When new committers has more experience of ShardingSphere, we can let them
to management their commit themselves.

------------------

Liang Zhang (John)
Apache Sharding-Sphere & Dubbo


zhaojun <[email protected]> 于2018年11月23日周五 下午1:01写道:

> I have changed mail setting :)
>
> Let us try to use `Squash and merge` & `Rebase and merge` , we’d like to
> share the experience each other.
>
>
> ------------------
> Zhao Jun
> Apache Sharding-Sphere & ServiceComb
>
>
> > On Nov 23, 2018, at 12:07 PM, 吴晟 Sheng Wu <[email protected]> wrote:
> >
> > I am not sure whether `rebase and merge` to avoid scenario(1) or not.
> `make commit log more clear` is what we try our best to do.
> >
> >
> > Also remind again, please change the mail name setting, to let others
> know who is talking :)
> >
> >
> > ------------------
> > Sheng Wu
> > Apache SkyWalking & Sharding-Sphere
> >
> >
> >
> >
> >
> >
> >
> > ------------------ Original ------------------
> > From:  "cherrylzhao"<[email protected]>;
> > Date:  Fri, Nov 23, 2018 11:59 AM
> > To:  "dev"<[email protected]>;
> >
> > Subject:  Re: Recommand to open `Squash and merge` as default merge
> inGitHubrepo
> >
> >
> >
> > Thanks a lot :)
> >
> > I have exactly understood what keep `squash` in default `commit merge`
> is available mean.
> > Also we should use rebase to make commit log more clear and reliable.
> >
> >
> >> On Nov 23, 2018, at 11:11 AM, 吴晟 Sheng Wu <[email protected]> wrote:
> >>
> >> I have no doubt our initial committers could do this well. But I
> suggest this because of new contributors. Let's think in this scenarios
> >>
> >>
> >> 1. A new contributor, and new to GitHub. We checkout his fork and send
> pull request. But he didn't bind this GitHub mail to his local git. Then if
> this PR merged, you will lose the visible way to track his contributions.
> GitHub will not show him in the contributor list.
> >> 2. A pull request is complex in some ways, and got reviewed/change
> requests by many times, so the PR author change codes a lot of times. Then
> how should we merge this PR? Ask him to rebase, or we don't merge? Or merge
> this by bring several dozens times commits, which will make our revalue
> more different.
> >>
> >>
> >> Both of these came from SkyWalking's experience, so, as soon as
> ShardingSphere joined the Incubator, I am sharing this for you. Also, you
> should know, this is the common ways in many Open Source projects.
> >>
> >>
> >> ------------------
> >> Sheng Wu
> >> Apache SkyWalking & Sharding-Sphere
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> ------------------ Original ------------------
> >> From:  "cherrylzhao"<[email protected]>;
> >> Date:  Fri, Nov 23, 2018 09:24 AM
> >> To:  "dev"<[email protected]>;
> >>
> >> Subject:  Re: Recommand to open `Squash and merge` as default merge in
> GitHubrepo
> >>
> >>
> >>
> >> In fact, we also use dev branch to add new feature like GitHub issue
> #1238.
> >> Usually we make issue id related with commit comment. `squash and
> merge` maybe not suitable for this scenario.
> >> I think we should make commit much smaller and organized by functions
> or modification purpose is better like Willem said.
> >> Also we can rebase the commit count if the content is not clear.
> >>
> >> On 2018/11/22 10:38:43, "[email protected]" <[email protected]> wrote:
> >>> Yes, right now ShardingShpere always use `dev` branch, and other
> branch>
> >>> just for temporary proposal. We can try to use `squash and merge`.>
> >>>
> >>> ------------------>
> >>>
> >>> Liang Zhang (John)>
> >>> Apache Sharding-Sphere & Dubbo>
> >>>
> >>>
> >>> Sheng Wu <[email protected]> 于2018年11月22日周四 下午3:19写道:>
> >>>
> >>>>> But if we have multiple branches and we need to cherry pick the
> patch>
> >>>>> between two different branch. It could be better if the commits is>
> >>>>> much smaller and organized by the functions or the modification>
> >>>>> purposes.>
> >>>>>
> >>>> Agree. But just fit for multiple branches in maintaining. So keep
> `squash`>
> >>>> in default, `commit merge` available is a good choice.>
> >>>>>
> >>>> For ShardingSphere, I don't see that is happening.>
> >>>>>
> >>>> Sheng Wu.>
> >>>>>
> >>>> On 2018/11/22 03:21:33, Willem Jiang <[email protected]> wrote:>
> >>>>> If the commits are only for fixing the PR review, I think 'sqash
> and>
> >>>>> merge' is good way to go.>
> >>>>> But if we have multiple branches and we need to cherry pick the
> patch>
> >>>>> between two different branch. It could be better if the commits is>
> >>>>> much smaller and organized by the functions or the modification>
> >>>>> purposes.>
> >>>>>>
> >>>>> Willem Jiang>
> >>>>>>
> >>>>> Twitter: willemjiang>
> >>>>> Weibo: 姜宁willem>
> >>>>>>
> >>>>> On Wed, Nov 21, 2018 at 11:02 PM 吴晟 Sheng Wu <[email protected]>>
> >>>> wrote:>
> >>>>>>>
> >>>>>> Hi, initial committer>
> >>>>>>>
> >>>>>>>
> >>>>>> I have known, we are going to move the repos. I think we could
> expect>
> >>>> ShardingSphere will have more and more contributors from out of
> initial>
> >>>> committer team. In our Apache releases, we need to provide
> CHANGELOGS, ref>
> >>>> SkyWalking's[1]. We should make commit logs matching the issues and>
> >>>> changelogs.>
> >>>>>>>
> >>>>>>>
> >>>>>> Also, in the future, we need to evaluate new committer and PPMC,>
> >>>> `squash and merge`makes your commits[2] list more reliable. Because
> in my>
> >>>> personal experiences, some one will submit a lot of commits in PR.
> Others>
> >>>> will rebase, especially the one has more open source experiences.
> Make>
> >>>> commits at least equal the number of PR merged.>
> >>>>>>>
> >>>>>>>
> >>>>>> Of course, this is only my suggestion.>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>> [1]
> https://github.com/apache/incubator-skywalking/blob/5.x/CHANGES.md>
> >>>>>> [2]>
> >>>>
> https://github.com/sharding-sphere/sharding-sphere/graphs/contributors>
> >>>>>>>
> >>>>>>>
> >>>>>> ------------------>
> >>>>>> Sheng Wu>
> >>>>>> Apache SkyWalking & Sharding-Sphere>
> >>>>>>
>
>

Reply via email to