On Fri, Sep 19, 2025 at 02:59:29AM +0200, Volker Diels-Grabsch wrote:
> Stefano Brivio wrote:
> > On Tue, 16 Sep 2025 21:21:11 +0200
> > Volker Diels-Grabsch <v@njh.eu> wrote:
> > 
> > > This incorporates all feedback of v5, and the commits are
> > > slightly reordered so the comment of the last commit makes
> > > more sense.
> > 
> > Applied, thanks to sticking to this! I hope this will finally
> > make the QEMU disconnect/reconnect behaviour robust enough.
> 
> As far as I can tell from my tests, locally as well as on production
> systems, this is pretty much the case.
> 
> Thanks a lot for the opportunity to contribute some helpful
> improvements to passt!
> 
> > A couple of notes for future changes:
> > 
> > - you can add this kind of message as cover letter instead, just
> >   git format-patch --cover-letter and git will format things as
> >   needed, and concatenate In-Reply-To: and References: email
> >   headers
> > 
> > - reporting the version number in every subject line might make
> >   reviewers' life marginally easier and it just takes a
> >   --subject-prefix="PATCH v6" (in this case) or even something
> >   like -v6 argument to git format-patch
> > 
> > - carry Reviewed-by: tags if you don't... change the change. For
> >   example, here, 4/5 was already reviewed by David. I added the
> >   tag back (and will always do anyway)
> > 
> > ...in any case, those are very minor details that don't really
> > cause me any trouble as a maintainer.
> 
> I believe these would be great additions to the upcoming
> CONTRIBUTING.md, especially for people like myself who are used to
> workflows like Gerrit or pull requests (a.k.a. merge requests) and had
> no prior experience with the kernel-style email based workflow.

CONTRIBUTING.md is now merged :).  I'm sure it will want some
revisions and extensions, but it's a start.

> (And moreover, the interactive interface of "git send-email" is far
> from self-explanatory and requires quite a lot of trial and error to
> get everything right, and usage mistakes are mostly impossible to fix,
> since the emails are then already sent.)

You're not wrong.  Fwiw, I've been pretty happy with git-publish [0]
which handles some of this workflow.  It may be more suitable for
regular rather than occasional contributors, but it might be worth a
look.

Admittedly, I'm not the best to judge its usefulness to someone new to
this, since I've been using the email workflow for 25+ years.

[0] https://github.com/stefanha/git-publish

-- 
David Gibson (he or they)	| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you, not the other way
				| around.
http://www.ozlabs.org/~dgibson