hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alan Gates <>
Subject Re: Reviews & commits (RTC/CTR), contributions, bylaws
Date Tue, 12 Apr 2016 15:32:14 GMT
I’m +1 on this change of allowing simple cleanup changes without requiring a full review.

But jumping to this fix obscures a bigger problem we have as a community.  This fix only works
for committers, not for non-committers who may also contribute such patches.  And it doesn’t
solve the situation for non-trivial patches.  We’re leaving potential contributions on the
floor and keeping people out of our community.  We need to solve this.

One thing I’ve been doing over the last few months is set up a filter in JIRA for components
that I know well (metastore, acid, etc.) and then put a recurring task in my task tracker
app to review a patch every day.  Realistically I manage 2-3 reviews a week, but that’s
1-2 more than I was doing before.  I encourage my fellow committers to find something that
works for them.  We need to improve the health of our community.


> On Apr 12, 2016, at 07:56, Lars Francke <> wrote:
> Thanks Thejas for the suggestion & others for jumping in. That seems fine
> for me. 2 days also seems good. Holidays are different in almost every
> country so I wouldn't exclude those.
> I have followed the procedure used for the last Bylaws change and created a
> new Wiki page here: <
>> .
> It includes this paragraph: "Minor issues (e.g. typos, code style issues,
> JavaDoc changes. At committer's discretion) can be committed after
> soliciting feedback/review on the mailing list and not receiving feedback
> within 2 days."
> I'm not a native speaker so feedback is welcome.
> I also fixed three typos in the Bylaws (and marked them as changed): <
> Once the discussion settles down I'll open a vote thread on the user@
> mailing list which requires a 2/3 majority of all active PMC members. I
> couldn't find a definition of "active" though.
> On Mon, Apr 11, 2016 at 10:26 PM, Thejas Nair <> wrote:
>> I agree we have a problem here. At least patches as small as this
>> shouldn't take too long to get reviewed.
>> Knox seems to consider a very large set of patches as being under CTR
>> process.
>> I think hive is very large and mature project that I would lean
>> towards RTC process for most issues. I think we can make an exception
>> for very minor patches such as fixing typos and and checkstyle issues.
>> Maybe the process can be to solicit reviews for such minor patches by
>> sending an email to dev@ list and if no response is seen in 2 days, go
>> ahead and commit it ?
>> On Mon, Apr 11, 2016 at 6:38 AM, Lars Francke <>
>> wrote:
>>> Hi,
>>> I've been a long-time contributor to Hive (5 or so years) and have been
>>> voted in as a committer and I'm very grateful for that. I also understand
>>> that my situation is different than most or lots of committers as I'm not
>>> working for one of the big companies (Facebook, Cloudera, Hortonworks
>> etc.)
>>> where you can just ask someone sitting next to you to do a review.
>>> I'd really like to contribute more than I do currently but the process of
>>> getting patches in is painful for me (and other 'outside' contributors)
>> as
>>> it is hard to get reviews & things committed. The nature of most of my
>>> patches is very minor[1] (fixing typos, checkstyle issues etc.) and I
>>> understand that these are not the most interesting patches to review and
>>> are easy to miss. I don't blame anyone for this situation as I totally
>>> understand it and have been on the other side of this for other projects.
>>> Is there anything we can do to make it easier for me and others like me
>> to
>>> contribute here? I absolutely see the value in having "cleaner" code and
>>> when done in small batches it's usually not very disruptive either.
>>> The bylaws currently require a +1 from a committer who has not authored
>> the
>>> patch. Knox for example has a different policy [2] where they distinguish
>>> between major features and minor things which can be committed freely.
>>> Hive could adopt something similar or like a middle ground. These are
>> just
>>> two suggestions:
>>> 1) Allow minor changes (up to the committers discretion) without
>> requiring
>>> an extra +1
>>> 2) Allow minor changes (up to the committers discretion) with Lazy
>> approval
>>> (i.e. wait 24 hours)
>>> Sorry for the long rant but I'd love some feedback on this and am looking
>>> forward to contributing more in the future.
>>> Cheers,
>>> Lars
>>> [1] e.g. <>
>>> [2] <

View raw message