openerp-community-reviewer team mailing list archive

[Guewen Baconnier <guewen.baconnier@xxxxxxxxxxxxxx>]

Hi,

On Fri, Dec 20, 2013 at 7:13 PM, Yannick Vaucher
<yannick.vaucher@xxxxxxxxxxxxxx> wrote:
> Hello dear community reviewers,
>
> Just to say few words about status on MP in launchpad.
>
> [...]
> Old MP without activity -> WIP
> Today, I took few time to change status of MP relatively old MP from "Needs
> review" to "Work in progress" in case there was a Need fixing that wasn't
> answer for more than 1 month.
>
> [...]

I guess everybody will agree with that.  I saw that Maxime put several
MPs to "Work in progress" today, thanks both for the cleaning.

>
> 2 or more Approves (review) before end of timer -> Approved (MP status)
> An other thing I did was to set few MP in Approve status as there was 2 or
> more approve but had to wait the minimum 5 days rule.
> That way we can return on the list of MP and see quickly what can be merged
> as time is up on the counter for waiting more eye balls. This could avoid
> some forgotten MP that could simply be merged. Of course, if in the mean
> time someone sees something to change. He can just change the MP status back
> to need reviews.
>
> Otherwise when merging a MP, I always takes the time to approve it first. So
> the approved revision of  the MP is written in it and it could help to see
> if some unasked extra code was added since it was approved.

I think it is a good idea, let see if we can be disciplined enough :-)

>
> Resubmit? Superseed?
> And tell me if I'm wrong, but the "Resubmit" type of review is there to ask
> the commiter to create a new MP like on an other branch because he forgot to
> aim and missed his target branch. I see the main use from commiter to say
> they resubmited and that previous reviews asking for "Needs fixing" are
> obsolete. Shouldn't it be a superseed to reset all reviewers? We can do
> superseed while keeping comments of previous branch. Last commits can still
> be seen in superseeded merge and the diff will be complete.

I'm not sure to understand what you mean. When you say "commiter", do
you mean "reviewee" (the one who propose a merge) actually?
I would use the word "commiter" for the one which has the rights to
commit on the main branch, as we are all here.
For the usage I could see for "Resubmit": wrong target branch,
conflicts. But from my researches, this is unclear to everybody, even
the Launchpad guys themselves I'm sure ;-)
The reviewee should not use this review type, only the resubmit action
if asked. I can't tell if in which case the supersede should be used
or not, is better or not...
My guess is that Launchpad has been widely developed for the internal
Canonical's workflows and not all the features are really usable for
everybody (try to use the blueprints, gasp).


>
> (Still wondering if my preview mail to community reviewers for accepting me
> in was received, anyway in short: glad to be part of it, thanks)
>

I think it was! You deserve it.

Thanks,

Happy seasons greetings and holidays!

Guewen