Mailing-List: contact dev-help@geode.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@geode.apache.org
From: Kenneth Howe <khowe@pivotal.io>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Mime-Version: 1.0 (Mac OS X Mail 10.3 \(3273\))
Subject: Re: [DISCUSS] Using Pull Requests over Review Board
Date: Wed, 14 Jun 2017 08:57:50 -0700
References: <CAA1wVxrdZhikQFTOBxUXa0vUBVcwCufazc20mTN6kT2n+mOVrw@mail.gmail.com>
 <d6be51b1-8b00-fd5e-963f-17a4850cb3ca@pivotal.io>
 <CAA1wVxoshxzmEGhbQ93ce0Vq5Y+BXQhp66d3W81brf93dAY1nQ@mail.gmail.com>
 <57c5d69b-0502-d85d-e5b1-b516ccc8b97b@pivotal.io>
To: dev@geode.apache.org
In-Reply-To: <57c5d69b-0502-d85d-e5b1-b516ccc8b97b@pivotal.io>
Message-Id: <9CDAFD60-2353-4883-A62E-F96CB2A610FF@pivotal.io>
archived-at: Wed, 14 Jun 2017 15:58:08 -0000

I have to agree with Bruce here. This is an open community, and as such, =
dictating whether or not we can use reviewboard for reviews works =
against the openness. Furthermore, reviewboard is NOT restricted to =
committers - I submitted many reviews using RB before becoming a =
committer. Once the reviews were completed satisfactorily it was a =
simple matter to do a single push to my GitHub fork and create a PR. My =
reviewers were familiar with the changes and pulled in my PR.

I personally feel RB offers better tools for doing a review than PR=E2=80=99=
s through GitHub. Others obviously lean the other way.

> On Jun 14, 2017, at 8:44 AM, Bruce Schuchardt <bschuchardt@pivotal.io> =
wrote:
>=20
> Having to deal with a github repo instead of branches off of the =
Apache repo is an additional burden on committers.
>=20
> I still vote -1 and don't see a lot of support for this idea.
>=20
>=20
> On 6/13/17 3:03 PM, Jacob Barrett wrote:
>> On Mon, Jun 12, 2017 at 7:57 AM Bruce Schuchardt =
<bschuchardt@pivotal.io>
>> wrote:
>>=20
>>> It places an unnecessary burden on committers
>>=20
>> Considering committers need to use PR to commit changes from =
non-committers
>> how does reducing the number of review systems increase the burden on
>> committers?
>>=20
>>=20
>>> and git history is the
>>> definitive record of changes to the code so github pr history isn't
>>> really very useful.
>>=20
>> All unsquashed commit history is preserved through a PR.
>>=20
>=20