On Fri, Aug 23, 2024 at 12:13:44PM -0700, Evan Silberman wrote:
> Walter Alejandro Iglesias <wai@roquesor.com> wrote:
> > [a bunch of stuff]
> 
> From what I have observed there's basically one reliable way to get
> one's patches into OpenBSD, as someone making occasional contributions
> to scratch an itch, without any previously-earned trust from the
> development team:
> 
> 1. Send a patch (inline!) that makes a single atomic improvement to
> something, with some explanation in the email body for why you think
> it's better.
> 
> 2. Wait for someone to take an interest, pinging if necessary not more
> than once a week or so.
> 
> 3. If someone has feedback, revise your patch in response to that
> feedback, or else explain why you prefer your approach.
> 
> Repeat all of the above until someone lands your patch or you get tired
> and do something else. It is true that sometimes it feels frustrating to
> work this way, especially in cases like yours where you have a bunch of
> work you've already done and you think all of it is good. But it's
> really the only way that I've seen most kinds of patches sent to tech@
> get reviewed.
> 
> It's certainly regrettable that these expectations aren't clearly
> communicated; OpenBSD doesn't have a ton of "onboarding process" or
> anything of the kind, presumably because it's an old project mostly
> maintained by a quite small team of people who mostly know each other
> and prefer to operate by convention and consensus than by establishing a
> bunch of bureaucracy and community norms. As you were more or less told
> some weeks ago, sending a bunch of patches that you want to see all at
> once linked in a zip file is far from preferred. You're now trying to
> land a single patch, which is an improvement, but you've resent it
> several times a day, in two different versions, which is not.
> 
> In the process you have evidently annoyed at least two of the developers
> with the most interest in email (generally) and thus mail(1)
> (specifically), plus probably a bunch of people who have started
> filtering you out without mentioning it, by acting entitled to the time,
> attention, and deference of people who have more of a stake in the
> project than you do.
> 
> If you reduce the heavy volume of pings and try to keep the grievance
> and entitlement out of your messages, you might get lucky and someone
> will still have the patience to work with you on the changes you want to
> see. If you don't get lucky, you have the source and a permissive
> license and you can enjoy your version of mail(1) in good conscience.
> 
> Regards,
> Evan
> 
> 

There are some little patches of mine on base (cwm) and ports (pop3d),
some of them sligtly modified by the team, some as is.  But there is a
difference between a patch intended to fix a bug or adding some option
to trying to improve a bunch of functionality of an application, which
requires an idea and prior planning.  I know it's a lot to ask, but if
you take a look at the evolution of my proposal from the beginning
you'll see that I took into account all the points you indicate.  And
when I deviated from the way of doing things here it was for a reason.

I never assumed that anyone would be interested in my proposal, not even
those involved in the development of smtpd(8).  In general, of course,
no one has any obligation to respond, but, intuitively, I understand and
assume that if someone sends me feedback, they first found a gap in
their workflow to attend to my proposal, took a look at what I did, and
after I respond with a modified diff according to their criticism, they
respond again letting me know if I understood what they asked me and
agree with my modification.  I don't trust those who limit themselves to
throw the bomb and disappear, nor someone who totally rewrite what I
wrote along a year in a minute, even if he can do it.  If this is the
way of doing things here, don't bother in *plunk* me, I *plunk* myself.


Thank you Evan for taking you time to respond me. :-)