[cctalk-request at classiccmp.org: confirm 38290c8a992491eda604beff5a06ff20cd7e85f5]

[Mouse]

>>> Can someone please fix the mailing list software?  This has been
>>> reported every once in a while by a bunch of people for over ten
>>> years.
>> Bounces aren't caused by the mailing list, they're caused by the
>> destination mail server.

Depends on how you view `cause'.

My own mailserver is one what I believe must be a very few that will
reject mail for, for exmaple, being marked as containing 8859-1 text
but including octets like 0x92 which do not occur in 8859-1 text.  Yet
such bounces are actually caused by the composing MUA's nonconformance,
with every intermediate mailhost that handled the mail complicit by
acquiescence; my mailserver causes the bounce only in the immediate
sense of being the one that blows the whistle and throws down the red
flag, the one that flags the offence for what it is.

> What I've been wondering for a while is the span of time over which
> the bounces are counted.

I haven't read the code for that logic in any mailing list manager (or
at least if I have I don't recall).  But...

> I can understand shutting a subscriber off for getting 10 bounces in
> as many minutes.  On the other hand if those 10 bounces are spread
> over two months, it seems rather severe.

It depends.  If those 10 bounces in as many minutes were for 10
consecutive messages, perhaps.  If they were for 10 messages mixed
randomly among 250 others that were delivered fine, it seems somewhat
excessive to me.

If I were writing such code, I would pay attention to not only count
and time but also to non-bounced list traffic.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse at rodents-montreal.org
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B