Re: [Evolution] [Fwd: You have been unsubscribed from the evolution-list mailing list]

[Patrick O'Callaghan <poc usb ve>]

On Fri, 2016-07-29 at 06:32 +0100, Steve T wrote:
> On Thu, 2016-07-28 at 22:10 +0100, Patrick O'Callaghan wrote:
> > On Thu, 2016-07-28 at 17:43 +0100, Steve T wrote:
> > > -------- Forwarded Message --------
> > > From: evolution-list-bounces gnome org
> > > To: stevetucknott yahoo co uk
> > > Subject: You have been unsubscribed from the evolution-list mailing
> > > list
> > > Date: Thu, 28 Jul 2016 16:26:29 +0000
> > >
> > > Bye!
> > >
> > >
> > > Received the above 'unsubscribed' message again today. I have had
> > > EMails from the Evolution list over the past few days, so I'm not
> > > sure
> > > the 'bounces' reason follows.
> > >
> > > Kind regards,
> > >
> > > Steve T
> > >
> > > PS I've resubscribed to send this
> >
> > What was the last message you received from the list before this
> > happened, and when did you receive it?
> >
> > poc
> Patrick,
> The last two mails were from yourself in reply to my posts with regard
> to the previous unsubscription issue. The two messages are dated/timed
> yesterday at 10:39 and 10:41.
> Steve T

The Mailman bounce engine accumulates a score against possible problem
addresses, and when it reaches a threshold it does the auto-
unsubscribe. Currently the threshold is 3.0, and scores can be either
0.5 (soft bounces, e.g. mailbox full) or 1 (hard bounces, e.g. user
unknown). Scores accumulate only once per day, no matter how many
bounces are detected, so even in the worst case it should take 3 days
for this to happen. However non-bouncing messages don't immediately
reset the score. Instead a tunable interval (currently 7 days) has to
go by with no further bounces, after which the score is regard as stale
and reset to 0. To put it another way: if you consistently get bounces
more often than once a week, you will eventually hit the threshold.

Obviously we can tweak these numbers (e.g. raise the threshold and
lower the freshness interval), but I'd like to understand what's going
on because in all likelihood it's going to keep happening, just less
often.

poc