Re: [Question] patch posting process

[Vadim Lomovtsev]

Hi Willy,

Thank you for your opinion, it's very helpful.

WBR,
Vadim

On Fri, Apr 06, 2018 at 09:21:46PM +0200, Willy Tarreau wrote:
> Hi Vadim,
> 
> On Fri, Apr 06, 2018 at 11:29:16AM -0700, Vadim Lomovtsev wrote:
> > Hi all,
> > 
> > I bring my Apologise for wasting your time, but ..
> 
> Questions about doing things right rarely are a waste of time if they save
> others from having to do useless work!
> 
> > May I ask for some clarification.. When we're speaking of 'posting patches shortly'
> > does it mean to send them in next few hours ?
> > Or would it be more acceptable to post one version per day
> > even for very small changes in between ?
> > 
> > Kernel posting guides says that one should wait for about a week for respond,
> > but in my case I've got feedback rather quickly (thanks a lot for that!)
> > and I'd assume that I can proceed with posting next version.
> > 
> > So, what is the proper approach here - should one wait day or two
> > before posting next version even if changes are very simple ?
> 
> Generally speaking, it's better to proceed ASAP. Reviewing patches requires
> some concentration and often some time to get into the context. Speaking for
> myself only, when I'm reviewing patches (I reserve time to do it), I prefer
> to get 3 round trips the same day than one per week and each time having to
> try to recall what it was about and what I proposed.
> 
> Also some people may only do that on spare time, week-ends or dedicated day
> in the week. If you sit on their e-mail for no reason, you expose yourself
> to the risk of having to wait for the next feedback. This is where the week
> comes from. Another nice side effect of the week delay is that some people
> send a first version for reviewing and figure by themselves that this
> version is bogus, then send a fixed version. That reduces the number of
> required work for reviewers.
> 
> On the other hand, it's not nice to rush quick updates without verifying
> that you properly addressed all reported points (addressed either in code
> or discussion). Thus my recommendation would be that if you can iterate
> one or two extra rounds the same day, that's generally much better. And
> in any case if the reviewer doesn't have more time to assign to you, he
> will switch to something else and you'll have to wait. Thus the good rule
> could be that ideally reviewers should not needlessly be waiting for you.
> 
> One important point however is *not* to send multiple versions of the
> same series without waiting for a review. Someone might already be reading
> your patchset and be pissed off by discovering he's reading outdated
> code. Reserve this for the cases where you've let a huge bug slip
> through.
> 
> Just my two cents, others will very likely have other advices.
> 
> Willy