On Mon, Nov 1, 2021 at 12:04 AM Vladimir Oltean <vladimir.olt...@nxp.com> wrote: > > On Sat, Oct 30, 2021 at 04:31:48PM +0300, Ramon Fried wrote: > > > > > > Ping for apply? > > > > > I'll get to the patches today. > > > > > thanks for waking me up :) > > > > > > Why don't you apply patches immediately after reviewing them? > > It wouldn't make a difference. you were all working on the same file > > in code, If I would have applied bin's patches before yours, then you > > would need to rebase. > > simple as that. Also the time when I'm applying and when I'm reviewing > > is also meaningless because this problem would occur anyway. > > I know, I was reacting to the 'thanks for waking me up' part. > Honestly I don't see the point of going through some patches, marking > them as reviewed, then waiting for 2 more weeks to apply them (or even > more in other cases). It slows down the process for a reason that I > can't see. > > Maybe the thinking is along the lines of "I won't apply patches that > haven't marinated for long enough". But my patches (the ones conflicting > with Bin's) still had build warnings even after marinating for close to > one month (fixes for which I've sent just now). During that time, I had > also piled up a few non-critical fixes to those patches, but I wasn't > going to modify the patch set already in flight anyway, for fear of > waiting again for who knows how long. Incremental fixups are fine and > surely are a valid way of doing development too, so that argument kinda > falls off the table. What's left? I do let patches marinate for a while, to give other people the opportunity to review, test and comment. After a few weeks, when the merge window is open, I pick-up all the patches and apply them in turn. That's my process, and I don't think there's anything wrong with it.
If you have a fix to a patch you already got reviewed-by, send the patch again, and add reviewed-by. That's the way it should work.
