← Back to team overview

openerp-community team mailing list archive

Re: New Community Organization for OpenERP

 

Op 22-11-12 09:58 schreef Alexandre Fayolle:
On mer. 21 nov. 2012 09:38:10 CET, Alexis de Lattre wrote:
Le 16/11/2012 16:41, Joël Grand-Guillaume a écrit :
For all of these projects, the rules we expect you to adhere to and
respect are:

  *

     Nobody merge his/her own work into the branch. Another member of
     the team must do it. Exceptions may be accepted if the merge
     proposal has been strongly approved by the rest of the team;

I agree on this, but I think it requires another exception : if you
are the main author of the module, then you don't need to go through a
merge proposal, you can push your code directly. For example, I am the
main author of the OpenERP-Asterisk connector ; if I apply your rule
to the Asterisk-OpenERP connector on Launchpad, that mean I would need
to try to find someone to review all my code on this project ; I'm
afraid that nobody will review it, because I am basically the only one
to master the code of this module !
Hello Alexis.

IMO, you should still try to go though a MP, and partners who are using
the module shoudl try hard to review the code. I agree that is no one
cares about a MP and does a review, we should have a way of forcing the
merge anyway, in order to avoid a lock in the situation.

Reviewing a MP does not require mastering the code of a module. I
recently reviewed code related to a specific format of 2D barcode, with
no knowledge of the format of the barcode and little knowledge of the
rest of the module. This did not prevent me from finding a bug in the
code and to suggest several improvements and cleanups. It also gave me
some knowledge of that piece of  code. Reviews are useful because they
enable non experts to get to know the parts of the module which are
evolving.

Hi Alexandre,

I agree with Joël. I see your points, but this is not a matter of "best possible code", but of "decent manners".

There is a saying: "You cannot look a given horse into the mouth". Meaning: when you are given a unique contribution, it is impolite to refuse it because you think it could and should have been done better. You should accept it and - when you have different standards than the giver - enhance its quality by either contributing bug reports or by contributing bug fix proposals. Reviewing and learning can be done afterwards as well.

Don't endorse mechanisms that enforce impoliteness.

Just my 2c,
--
Pieter J. Kersten
EduSense B.V.


Follow ups

References