The future of commit access policy for core Firefox

Ehsan Akhgari ehsan.akhgari at gmail.com
Sun Mar 12 20:40:43 UTC 2017


On 2017-03-11 9:23 AM, smaug via governance wrote:
> On 03/11/2017 08:23 AM, Nicholas Nethercote wrote:
>> On Sat, Mar 11, 2017 at 2:23 PM, smaug via governance <
>> governance at lists.mozilla.org> wrote:
>>>
>>> I'd be ok to do a quick r+ if interdiff was working well.
>>
>> Depending on the relative timezones of the reviewer and reviewee, that
>> could delay landing by 24 hours or even a whole weekend.
>>
> The final r+, if it is just cosmetic changes wouldn't need to be done by
> the same reviewer.

OK, but what is the exact time zone efficient way to ensure that no huge
amount of delay is added for the turn around of this final review round?

Let's be realistic.  In practice, adding one extra person in the process
of landing of the patches that currently land with r+-with-nits *will*
slow them down.  And we should expect that it will slow them down by
factors such as time zones, people missing bugmail, etc, all of the
reasons why currently one review can end up taking a week, it may end up
being that final round of review after this proposed change.

And I still don't understand what the proposal means with rebases in
practice.  What if, after automation tries to land your change after you
got your final r+ the final rebase fails and you need to do a manual
rebase?  Do you need to get another r+?  What happens if you're touching
one of our giant popular headers and someone else manages to land a
conflict while your reviewer gets back to you?

> Perhaps we shouldn't even call the last step a review. It would be
> "ok-to-land".
> r+ without asking any changes would implicitly contain that "ok-to-land".
> (if rebasing causes some changes, that would then need explicit
> "ok-to-land")

Are you suggesting a change to the nature of the review process for the
last step with the rename?



More information about the firefox-dev mailing list