[Drbd-dev] [PATCH 00/18] use semicolons rather than commas to separate statements

[Joe Perches]

On Sat, 2020-10-03 at 15:31 -0400, Konstantin Ryabitsev wrote:
> On Sat, Oct 03, 2020 at 09:18:51PM +0200, Julia Lawall wrote:
> > > > There seems to be some mismatch between b4's use of the
> > > > cover letter to a patch series and what maintainers that
> > > > apply a subset of the patches in the patch series.
> > > > 
> > > > The merge description shows the entire patch series as
> > > > applied, but the actual merge is only a subset of the
> > > > series.
> > > > 
> > > > Can this be improved in b4?
> > > 
> > > So, the following logic should be applied:
> > > 
> > > - if the entire series was applied, reply to 0/n
> > > - if a subset only is applied, reply to each n/n of the patch that was
> > >   cherry-picked out of the series
> > > 
> > > Is that an accurate summary?
> > 
> > That sounds good.
> 
> I'm worried that this can get unwieldy for series of 50 patches where 49 
> got applied. Would the following be better:
> 
> -----
> From: ...
> To: ...
> Subject: Re: [PATCH 00/18] use semicolons...
> 
> On Sun...
> > These patches...
> > 
> > [...]
> 
> A subset of these patches was applied to
> 
>   https://...
> 
> Thanks!
> 
> [5/18] regmap: debugfs:
>        commit:
> 
> (etc)
> -----
> 
> In other words, we:
> 
> - specifically say that it's a subset
> - instead of just enumerating the number of patches that were applied, 
>   as is currently the case ([1/1]) we list the exact numbers out of the 
>   posted series (e.g. [5/18])
> 
> I think this is a better solution than potentially flooding everyone 
> with 49 emails.

I think it would be better to reply individually as
the likelihood that the maintainer skips just a few
patches of a large series is relatively low.

It's more likely for a treewide or multi-subsystem
patch set for a maintainer to apply just a single one
or a selected few of the patches and individual
replies make it much easier to determine which ones
were applied.

thanks, Joe