Code review required for commits - Discuss.

[Matthieu Patou]

>> Of course I do understand the joke but in every joke there is a part
>> of truth, so what about people in .au or .us ? Should I knock on the
>> door of Jeremy to get a review ?
> You do know where I live :-).
Yeah prepare for a year long of a "a review or a trick" :-)
>> I don't think it's being fair to the developer to ask him to nag
>> reviewers in order to get a review. I do think that within 3 days a
>> reviewer should pop up to say: I want to review this and then maybe
>> the reviewer gets a bit more time to do the review but I still do
>> think (also) that it should be limited so that we keep on the
>> dynamic and review never get stuck for weeks as it has been the case
>> in the past for some voluntary reviews.
> Look at the things that have been stuck for weeks (and I'm
> mainly thinking of poor ddis's async RPC rewrite).
>
> They're mainly *big* changes that require lots of work to
> review.
>
> But you know what, they're the things that *should* be
> reviewed more thoroughly.
>
> I haven't seen the case where a simple patch didn't get
> a prompt ack.
I've got two examples of things that comes to my mind:

* dirsync control implementation
* pseudo-backlink

I took me ages to get a review on both (2/3 weeks), and I'm sure if I 
dig in my memory there is more like that.

Matthieu

-- 
Matthieu Patou
Samba Team
http://samba.org