Gateway to python-list is generating bounce messages.

Steven D'Aprano steven at REMOVE.THIS.cybersource.com.au
Thu Sep 11 05:10:00 EDT 2008


On Thu, 11 Sep 2008 15:25:57 +1000, Ben Finney wrote:

>> > The bounce messages are sent to you because you sent the original.
>> 
>> Wrong.  I didn't send _any_ e-mail.  Why should I get bounce messages?
> 
> You asked for email to be sent, by sending a Usenet post to
> comp.lang.python. That's what a news-to-mail gateway does.

I wasn't aware that comp.lang.python was a news-to-mail gateway. How can 
one tell the difference between news groups that use a news-to-mail 
gateway, and news groups that don't?

Posting to a public newsgroup is obviously giving consent to forward that 
news posting to news clients. Given the ability to opt-out with the X-No-
Archive header, there may even be an implied consent to allow archiving. 
But I don't believe there is any such implied consent to format-shift 
news messages to email.

In practice, I couldn't care less what format my news postings are 
converted to, so long as it is invisible and transparent to me. If 
somebody wants to build a news-to-carved-in-giant-stone-tablets gateway, 
I don't care, so long as I don't get giant stone tablets aren't dumped in 
my front yard.

Nor do I believe that by posting a news message I've automatically 
consented to receive email messages. To imply that the one implies the 
other is equivalent to arguing that because I've written a letter to the 
editor of a newspaper, I therefore must accept private correspondence 
from any person or corporation that has a subscription to that newspaper.

(In practice, I don't mind human-generated email messages, but not 
automatic messages. If you can turn my munged email address into a real 
email address, I probably won't mind you emailing me. Don't abuse the 
privilege.)


[...]
> I sympathise completely with your irritation at receiving bounce
> messages from poorly-configured software, but the solution is not to
> break the news-to-mail gateway.
>
> The correct solution is to unsubscribe the badly-behaving address from
> the mailing list, and refuse re-subscription from that address without
> assurance that the bad behaviour has ceased.

The problem with that "correct solution" is that the party who suffers 
isn't in a position to correct the problem, and the party who can correct 
the problem has little incentive to do anything about it. That makes the 
solution ineffective and therefore anything but "correct".

I hope the person running the mailing list does do the right thing, but 
if he or she does, it will be an accident of policy or personality, and 
not because the system is robust and self-corrects errors. To quote 
Dennis Lee Bieber:

"The bounce/ooo-reply is sent to the message author, not to any 
intermediate host(s). After all, on that end, it's normal email failure 
response -- notify the author of the message. It doesn't matter that the 
original message was posted on a Usenet newsgroup if that group is 
automatically relayed to members of a mailing list."

But that's wrong: it *shouldn't* be an normal email failure response, 
because the message author is in no position to do anything about it 
except to cease posting. That's a problem with all mailing lists (that I 
know of).



-- 
Steven



More information about the Python-list mailing list